Hmmm. This seems to be a recurring issue with newbies.
There is nothing like slowing down and figuring out a game. In fact, the longer I have worked on a game...the more I have found. Embarrassingly enough, all my TASes have been submitted with flaws...not because I didn't take a great deal of time to figure it out, but because I didn't do enough testing to find things that were not so obvious. Simple platforming games, like this one, still has challenges to get through. Its no all about holding "RIGHT". Spikestuff clearly made that known with his effort. Experience pays off, but if you have none...you have to start with the right approach. SLOW DOWN and stop racing to do something. I made this mistake with my very first submission and FAILED.
I think you'll be surprised how rewarding it is to put a decent amount of time into a TAS. Then length of a game certainly doesn't mandate a certain amount of time. If this wasn't so, then the Gameboy version of Tetris (B Game), wouldn't be in the MILLIONS of re-recording by a BOT. When this TAS was done, it cut around 6 seconds from the previous version.
Check out Mr. Wint's TAS:
http://tasvideos.org/4536S.html
If your TAS does get accepted...you take the chance of increasing the visibility of your submission and creating a challenge for others to obsolete yours.