I ain't no mathematician for sure, but reading the TM comments on the site in general it seem to me that TM uses the ball flight to calculate the variables that it cannot measure: as far as I know these variables make up what is now known as the "top vector". This fact coupled with the incredible control of ball flight, inc. trajectory, of the tour pros could explain this bizarre consistency (0.3* club to club).
For those who find this a bit cryptic, here it is in clear English: if the ball flight is X then it follows that AoA + Face must have been Y.
As Brian explained in his video about TM measurement true face angle cannot be measured because even with 100,000 fps cameras the measurable face angle at impact is not the true face angle relating to the D-plane.
So all in all we're looking at numbers which may (or may not be) MORE accurate than truly measured numbers.
PS is the difference between TM and FS PURELY the algorithms/formulas?