Wind tunnel testing measures drag which can be used to calculate power requirements. Time savings at a given power are calculated from the control which is usually road bike with 32 spoke box sections and a road helmet. Bikeradar did a
test last year which showed how everything was done and some basic calculations.
Figures on the internet usually quote time savings for a 40km TT. Most of the research is done on equipment and properly positioned riders - position is more important than the gear - your body being the big un-aero thing on top - so there is even more time to be saved there.
No idea about moving from MTB to road....slicks should make a huge improvement as the frame isn't as significant and aero trumps weight (if you can get into a decent position on it).
Aero bars are one of the cheapest and most cost effective upgrades as they put you in a much more aerodynamic position. The bikeradar wind tunnel test shows a saving of nearly 2 minutes over 40km....this sounds about right as a good rider managed to break his tribars on the way to the start at a club 10 recently...he went from a consistent 22:40 to 23:30 when forced to use the drop parts of his bars.
Helmets are a minefield because of all the different shapes and sizes of something which is really exposed to the wind, plus whether they sit properly on the back.
Various tests suggest you could save anything from 30-90 seconds over 40km.
Zipp did some testing for their wheels against a Mavic Ksyrium control:
404s (58mm deep) - 62 seconds // 20W
808s (82mm deep)- 72 seconds // 24W
404F + 808R - 64 seconds // 21 W
808 + 900 (disc) - 82 seconds // 27W
You can see from this that the disc is worth about 10s // 3W over an 808 rear, but note also what happens when changing a 404 rear to an 808 - only 2 seconds // 1W....the front wheel is the most important as it is what breaks the air in the first place.
Can't find any data on clothing, but I know I've seen some before. The story here is that flapping is bad. Some even suggest that gloves slow us down!