I'd like to thank the following people:

The SDA staff - seemingly increasing by the day, this can only be a good thing!
Nate - for various feats of technical wizardry (including developing Anri-chan, which was used to encode this run)

My girlfriend Emma - allowed me to practice on her Sonic Gems Collection and record with her Dazzle, I actually got an 11:40 at her place that I nearly accepted. I love you :)
VecGun - for two (or maybe three?) previous runs
mmbossman - for a well-made TAS, and in particular one trick that came out of it: shooting up-right as you land to reset your horizontal speed, which found a lot of use in Day 3

Vectorman is a game that I just can't seem to stay away from. A few years ago I picked up Sonic Gems Collection, and indeed ran Vectorman 2, but I kinda wrote Vectorman off for running since a) there was already a run up, and b) I didn't quite like the physics as much (V2 doesn't have those insanely annoying turrets, for example).

Now there's been a handful of Vectorman runs by this stage (either two or three) and I've had the most recent of them on my hard drive for some time. Even looking at it casually I could tell it was improvable, and I kind of toyed with improving it by 30 seconds to a minute.

And then came Classic Games Done Quick.

Even before this I had already been planning to make the trip for MAGFest, but now I had some runs to practice for. Undoubtedly I would pick up Sonic 3, and Sonic 1 was slated to be a race with stanski (who eventually was unable to participate, meaning I had to take Sonic 2 as well). I also picked up DKC3 (which, thankfully, I did not have to run) and was also considered the best candidate to run the two Vectorman games. That meant finally coming back to this one and seeing what I was capable of.

Even before MAGFest, I hit a 12:06, which led me to believe I could beat the existing speedrun in the live attempt at CGDQ. (I didn't :( ) While staying with Emma I managed 11:40, then once I got home I made the decision to change a couple of strategies and hit a very strong 10:56. I sat on this run for a while and eventually decided to send it in. However, shortly after I sent it in, Mike went to Europe for a month, which meant the queue remained stagnant. After a while spent experimenting with a Hard low% category (which I later abandoned), I took what I learned from that back to this run and brought it down to this mark of 10:24. Much of the time saved comes from skipping both extra health pieces that I picked up in all the previous mentioned attempts (Days 6 and 8).

In the end, this run beat its predecessor by almost four minutes, and beat every single individual level of the previous run. I still estimate there's about 10 seconds of improvement scattered throughout, but you'd need immense luck to get it.

Individual level comments:

Day 1: Terraport (0:54) [-0:04]
The route here remains largely unchanged. At the end though, I elect not to take the bolo shooter and just go with regular shots instead. I actually succeed in doing the boss in two and a half passes, almost as fast as VecGun with the bolo! To be fair, though, I do have the speed advantage... but even without it three rounds with the regular gun is certainly not slower than two rounds with the bolo, maybe a fraction faster.

Day 2: Metalhead (0:07) [-0:06]
Not sure what took VecGun so long here... you're going to take two damage anyway, just stand in place and shoot. Of note is that it's easier to get fast times here than in the NTSC version because there's no limit on how fast you can shoot, so I'm just mashing both buttons here. Without taking increased reaction time into account, this is the one major difference that favours PAL over NTSC, and it only accounts for a few seconds at best.

Day 3: Tidal Surge (0:59) [-0:21]
This level benefits the most from the TAS. Firstly it brought about the complete route change, moving towards taking the top route the majority of the time, which proved much quicker. Secondly, the momentum-conserving technique (shooting up-right as you make contact with the ground) discovered by mmbossman enables you to get lots of speed even on a short platform like the ones the turrets stand on, which enables you to make even longer jumps and cover more ground in the air.
Got knocked down to one health early, though at least the bounce was favourable. After that I was a bit disoriented and jump-boosted an enemy I don't normally kill that way and picked up a life under x3. No complaints. I have yet to beat :59 so far, a few times I've come very close but never beaten it, so all that matters nought.

Day 4: Absolute Zero (1:00) [-0:24]
Pretty clean run of the level. There's a tricky jump at 6:50; because of where the first Fish Morph is placed, if you get a jump from the very edge of the platform next to it, you can pick it up and carry your momentum upwards, which saves a couple of seconds getting to the second one.
Bosses played really nice. Borealis in particular stood straight up... you lose two or three seconds if he never stands up, and it's also possible he'll stand up somewhere in between.

Day 5: Arctic Ridge (0:36) [-0:13]
Lifted this path straight off the TAS. The previous run took a really slow path, basically going down the pits and then up out of them, whereas I just cannon straight through. Had a bit of luck at the first gap, though to be honest it was about time, that jump had been giving me trouble for ages. The TAS simply did a perfect jump-boost and landed in exactly the right spot at the end, but I'm not quite crazy enough to attempt that.

Day 6: Bamboo Mill (1:04) [-0:25]
Apart from one of the bugs randomly bombing me, the level was pretty clean. Annoyingly jumped straight over a life piece though, which could have been useful to have.
Major strategy change: no stopping for items here. I first omitted the spread shot, thanks to practicing on hard low% which didn't permit me to pick it up. Having had so many attempts at that fail to Warhead, I toyed with the idea of picking up no extra health at all on normal difficulty. When my first attempt reached level 14, I knew it was well within my reach.
Clockwork kinda sucked, got hit twice in a row, but I lived through it and it was reasonably quick.

Day 7: Rock n Roller (0:06) [-0:03]
Straightforward enough. For such a short level the time here can be wildly variable due to luck, but the right hand played nice and let me hit it a lot.

Day 8: Death Alley (0:27) [-0:10]
Slightly missed the ridiculous boost-into-turret at the start, the way that's supposed to work is that you boost just a bit later and hit the underside of the platform the turret is attached to. The problem is that normally a) you boost too late and the turret damages you and knocks you to the lower level as it dies, or b) you boost too early and this happens. This is definitely preferable though.
This is the other level that significantly benefits from not picking up the extra health; I estimate the difference is about six seconds here and 6-8 on Bamboo Mill.

Day 9: Hydroponic Lab (0:39) [-0:21]
I put a ton of time into this level, because it was the level I felt was done worst in the previous run, and in many ways was the catalyst for me attempting to improve it myself. Additionally, this level is actually substantially harder in the PAL version: there's four or five streams of bugs that, in the NTSC version, do not exist. I can only assume they were added to counteract the game being slower, or something. :/
I had a really precise route for this level, so for example sometimes you'll see me not shoot the fire extinguishers and they'll jump right over me as I approach them. I was really unlucky with health drops here: the one that I did get I jumped right over, again.
The boss requires a point of damage to do efficiently (which I didn't, I probably lost 1-2 seconds on it) which left me with just one health for Superstructure. Given Stayin' Alive requires me to have at least three health to do efficiently, this is not a good place to be.

Day 10: Superstructure (0:37) [-0:06]
Not my best here, but under the circumstances I'll take it.
This is a level that relies on the turrets playing nice. If they don't, you're screwed, plain and simple.
Again, a stream of bugs is added that didn't exist in the NTSC version, which makes the jump over a turret at 3:15 a pretty nerve-wracking one, especially if it moves around a lot.
Then we come to the Jet Morph section, where you might be invincible but the turrets can -still- screw you over! If they get a shot out on you, you get thrown backwards a little. The outcome I got was pretty good, any run that gets to the end of the second tier there is pretty good. I did kinda stuff up with my angles at the end and I might just have squeezed past that turret if I'd got it right. On saying that, though, I managed to pick up a third point of health off the very last enemy in the stage, which I was definitely happy with.

Day 11: Stayin' Alive (0:09) [-0:02]
Another button masher, but one which requires good aim. I tried to herd the tornadoes out of the way here, but one of them got me costing me a couple of seconds (you can easily get 6 or 7 if you get an uninterrupted shot at the weak point).

Day 12: Nightscape (0:30) [-0:19]
This is another level I learned very thoroughly because the previous route went way out of its way for a Jet Morph, which is a monumental waste of time since a) it takes a number of seconds to get in the first place, b) it's actually faster to fall than to use Jet Morph downwards and c) the enemy patterns are no more difficult than any other level. The main problem is, again, a couple of turrets - some of them that are perched at the top of ascents will take the opportunity to shoot diagonally down at you before you can boost them to death. That didn't happen here, though, which was lucky because I got knocked all the way back to one health after climbing back to four! Unlucky to be double-hit by one enemy (the boost was intended to kill it) and a flat-out mistake at the final ascent probably also cost me a second (29 is my best here).

Day 13: Dark Ruins (0:35) [-0:13]
This level was all about making the tough decisions. In preparation for my first revision of the improvement I decided to cut out the Max Health in this level because the detour was just too big. Originally I substituted it for the one in Worldlink Center, but I ended up skipping that one too, and then all the others, which made Warhead a risky proposition indeed. Still, in the name of speed I took the risk.
The second big decision was dropping the invincibility. The invincibility is right in the way, only costs a second or two to pick up and grants me a perfect rest of the level, with no strings attached. However, watching my first two runs back it turned out there weren't as many annoying bugs as I thought, so I made the decision not to pick it up and learn the last part of the level more precisely. It worked a treat: the only enemy of any consequence is the heavy right at the end, and I plowed right through him.
This level has had a habit of screwing me over in the past with random turret shots, and on one health I was definitely tense. But I managed to dodge the shot at 7:44, and everything else went smoothly. Tied my best time on this level.

Day 14: Underground Vault (0:56) [-0:24]
This level was not designed to be speedrun, there's nothing else to it. Made up a strategy for the Ascent of Doom on the spot which seemed to work pretty well, then got hit by the turret as I was delivering the final blow to it. Not sure how that happened, cost me a second.
I then pulled off a perfect Spider, a quick Piranha and a very ballsy Monkey fight - I picked up the extra health and launched straight into a jump-boost damage strategy. Took an unnecessary hit, but got the job done.

Day 15: Worldlink Center (1:16) [-0:21]
Some enemies in this level can be particularly annoying, but I was very lucky and didn't have any problems, even the turret/fire extinguisher combo played nice! Equalled my best time here, some of the turrets shot me in Jet Morph mode so I might be able to squeeze another second, but this close to the end I'm very happy with this.

Day 16: Twist and Shout (0:29) [-0:20]
Average fight. Warhead takes three well-placed jump boosts to kill and given you get ages to set up your first attack, getting two consecutive hits is usually no problem, as was the case here. It was getting back in position for the third hit that was the time sink, my best time here is 25 and I probably would have hit it with perfect luck on the third hit.

TOTAL TIME: 10:24 [-3:52]

There's some improvement that could still occur (better luck in day 16, for instance), but I doubt I would reasonably squeeze more than a few seconds out of it at this point, given the amount of luck I had throughout the run. I worked out that the sum of my best level times is about 10:01, so that's an indication that the theoretical limit might be closer to 10 minutes with insane amounts of luck. For additional reference, the sum of level times in the published TAS is 8:58.

Enjoy the run!