The TCAS data suggests he was doing 140-150 knots on descent from 10000', then accelerated to around 200 knots on 4.5 mile final. A search for GA-8 VNE suggests it is 185 kt.
The distances are short enough that an error of 1/2 mile one way at the beginning and 1/2 mile the other at the end brings the required speed down to a more plausible 120kt. But if you need a margin of error to make it fit, assuming accuracy in rest of the analysis doesn't make sense.
I found this interesting link about TCAS accuracy in turns:
http://www.tc.faa.gov/its/worldpac/techrpt/tctn099.pdf
I don't know how relevant it is to current systems. But the entire "near miss" occurred while the jet was turning. Before the turn, XGA was slightly behind and around 700 feet higher. When the jet straightens on upwind, XGA suddenly appears south of the airfield (where he said he was going to go when asked to go around)
During the turn, TCAS says he was doing 200 knots and almost ran into the back of the jet (0.1 mile separation). It makes a lot more sense if the TCAS data is unreliable in the turn.