Karrank is correct- the data on speeds of a/c type for the route from the feeder to landing are defined offline by ASA's TMA gurus and are updated if they find regular inconsistencies.
TAS, SHMAZ. Put 120k as a TAS in a 738 plan and you'll more than likely still cross the feeder at the same time if you had the correct TAS in the plan. The fact is if Maestro inputs that you'll cross narelle at 34 and our Bearing and Range Line says 26 and we allow for a couple of minutes for reduced speed on descent, we manually input the revised untouched revised time of 28 and that's what goes in the system. We check every time and update as required. Once locked in, Maestro CANNOT change that time, it simply calculates the delay required to get you evenly spread at the field.
It is surprising how much controllers are focusing on TAS in this thread. The only time this may affect your slot is if things are so busy and delays are extensive, controllers may abandon revising fix times and say stuff it, we'll go with what the system says.
Most of the time the times are set by us based on our estimates from the radar data, expected time loss on descent, crappy or not so crappy descent expected speeds. Patterns are also taken into account eg. we expect the cowboys in the westwinds will go like the clappers as long as they can, Qf can lose the standard two minutes plus an extra couple for the bean counters and some carriers like PAL want to be at A090 by 65DME so we add 5 minutes for them. Pure difined data does not work.
Cheers,
R-S.