I quite liked that. The height is an issue on that last flip like Ray said. Also, It wouldnt be as much as a problem if he didnt stay at relatively the same height the whole time. Try to make it a more gradual increase like a saturation graph. (Which just means ease the height so he starts off going up faster, and then slows down) and then doing the opposite of that when he lands, start slow going down and then faster. Another thing that looked a little weird was the free arm on that back-handspring, looked a little shakey.
Raymond wrote: Remember that nobody can just speed up or slow down after they are already mid-air.
While this is normally true, some of this does depend on body position. Since the body position doesnt do anything to cause more rotation the rotation shouldnt go faster. So in this case that is correct. However, when you do a back flip in the tucked position going from a long body and extending upwards on the block, to the short body (tucked) that causes rotation. Going from a long body to short body it the only way to go faster and the opposite is true for going slower. This is because there isnt any external force being applied to the figure, so in order to spin yourself you have to create an internal torque by distorting your body position. For flipping you move your arms and legs to create the torque and for spinning, you usually only use your arms, however you can use your legs. You can also use your head for rotation, but using it for anything other than a whip is bad form.
(an example of using your legs to help spin)
Sorry for the long response. I hope that I clarified, helped!