I don't see any other submissions from this user, so I'll assume it's their first.
To give you some background, I've been dealing with this particular game, in absence of the rest of the DKC GBA trilogy, for over 10 years. I'm in the process of making two TASes of it: one any% with the goal of the simple beating of K. Rool the first time, and 103%, the (without the TUFFER code) completion of everything.
There are some major setbacks to this run. First, I'll go over the rerecord count. I notice the rerecord count is extremely low for the amount of time the run lasts. Usually, when I TAS a game, I make about 7,000 rerecords every ten minutes. You made approximately 18,000 rerecords in this 2 hour 20 minute run. For me, 18,000 rerecords would be about 24 minutes of gameplay by that math. Especially for a game as complex as this one, the amount of rerecords should be very highly considered. The amount of routing in this game that I alone have to do for 103% is massive still, so really I'd probably say double my usual rerecord count, and expect around 14,000 rerecords per ten minutes, so your run would really equate to about 12 minutes of gameplay in that sense.
Secondly, the version. Usually, if a USA/Japan version of a game exists, one of either of those two versions are preferred over the European, since the European versions of games usually have some major flaws. There are some rare exceptions I suppose where the European version of a game may be faster by chance, so it's a case-by-case basis. In this case, the European version of Donkey Kong Country 3 for the GBA is far slower than the USA version. In the USA version, at the time any flagpole is touched, you can skip the flagpole animation by pressing Start+Select (if it's a first-time-in-a-level situation, this only works when you touch the flagpole). In the European version, this is not the case; that trick simply cannot be done. It is also absent of the teamthrow door glitch, and, according to Avanor_, the newly discovered time-stop glitch is impossible in the European version as well. So the obvious optimality is in the "USA, Australia" version.
Third, the optimality. I'd say this is the worst problem with this TAS, and evidence of the suboptimality of the TAS is shown in the last two points I made about rerecord count and version as well. While some parts of this run were entertaining, and many of the bonuses were very well-done in comparison, I saw not just a few, but several parts of the run that were pretty obviously (to me) suboptimal. Even without taking into account the many skips that weren't done in many parts of this run that would've saved massive amounts of time, there were portions of the run where you were just wasting a little time, presumably without knowing it. For instance, yes, you did do a lot of speed boosts with Dixie which did save a lot of time, but did you know that every time you press B with Dixie and she smacks her hair, it is significantly slower than regular running? You did this at the beginning of a lot of levels and also during a lot of levels for no real reason, perhaps without knowing of the consequences of this. I could name a lot more specific mistakes, but that was one that really jumped out to me.
The run clearly was not tested very well for optimality. I even saw that at the very beginning of the run, SpectrumDisaster did the game's built-in restart sequence by mistake without realizing it, while mashing random buttons to get past the "Press any button to continue" splash screen. (The restart sequence is done by pressing A+B+Start+Select) This obviously wasted quite a bit of time, and I feel like someone trying to optimize a TAS really should have been able to realize that they did that. Speaking of button-mashing: I noticed that instead of waiting for just the right frame to press buttons (such as in text-skipping, etc.) you used the autofire sequence. If the autofire sequence hits at the wrong times, it can waste a couple frames every time you do it, which is very clearly suboptimal.
Besides even all the things I just mentioned, if you know the game, (or maybe even if you don't), you'll see that the general manner of play is often sloppy or otherwise much less optimal than it should be. The routing needs a lot more work than what was put in as well.
Fourth, the category. The category is "105%". A lot of people have questioned what this actually means. So, we have the two main categories of the game: any% and 103% as mentioned by me above. But 105% is slightly different from 103% because 105% is based off of an in-game cheat code ("TUFFER") that makes the game "harder", removing DK barrels and save barrels, and also gives you an extra 2% when absolutely everything is completed, rather than just 103%. In the speedrunning community, this has been considered to be a separate category from 103%, and 105% is not considered "more 100%ish" than 103% in any sort of way, and is thus not a replacement of 103% runs. 103% is the true "100%" category, so I'd say that 105% is like a "side-category". The fact that one of these "side-categories" was submitted before any main categories were ever submitted is definitely troubling, since there's nothing main here to compare it to.
A separate problem with this side-category is that it relies on a cheat code to be performed at all. I'm not even sure that that is allowed here on TASVideos at all, though it isn't really considered cheating in this case.
Now I will say that you did make a good record with this run, beating the world record of this category by over 30 minutes, and also beating my 103% world record by 19 minutes (though that is considered a different category). In a general sense, I'd say that you did a decent job with this in comparison to regular speedruns (though there is a ton of room for improvement even in that sense).
In conclusion, I voted no in terms of entertainment value, and I highly suggest to the judges not to accept this run. Nothing personal, but just I wouldn't if I was judging it. First, it uses a bad version of the game. Second, the rerecord count does not show signs that the user spent a significant enough amount of time testing the game and producing this movie. Third, the run is clearly suboptimal and sloppy for a TAS. Fourth, the category is questionable.
Finally, a note to the TASer. I know my statements are harsh, and others after me might even be more harsh, but don't take it personally. I'm not saying that you as a person or your ability to do work are any less valuable. You are apparently a beginner. I was there once too. The thing is that the TASVideos community puts a lot of value on perfection, and if a TAS submitted here is perceived as less good than it could be in any way, especially when it's more obvious like in this case, reception is often harsh in this manner. I encourage you, if you want to continue your experience with tool-assisted speedrunning, to take more time to learn about TASing and take an exceptional amount of time learning about your games before making TASes of them and submitting them. Learn all about them. Make sure that everything you do couldn't be any faster; any more entertaining. Make absolutely sure that in any particular instance of the run, there isn't a part where you may have screwed up in some way, or did something even slightly slower than it could be. Really check. REALLY dig into the game, as if you were trying to find a small marble inside the dirt in an extremely wide grass field, and as if your life depended on finding it. This kind of precision is what makes great TASes, which are the kinds of TASes we want here.
