Okay, vehicle speeds 184.108.40.206.0.0.1
Basically, this is the most sensitive topic on the wiki purely because of how controversial it is...and the different methods we've encountered, from Broughys debatable 'measure distant divide by time' (which, as covered in my forum post linked above, doesn't satisfy me, as I have a hard time trusting the in-game distance values, let alone those displayed in the creator or on the GPS, of which he relies on. I also cringe at his averages, calculated on road surfaces that aren't truly flat and constantly give or take speed off you whenever you hit an invisible bump), to the highly infamous first person speedometers.
Truth is, neither of these work. Speedos had the potential to work, however the poor mapping of the textures and needles doesn't satisfy the value displayed, I.e, the needles DO use the real speed as per the file data, but the way the textures are aligned result in inconsistencies. The highly unreliable distances mean D/T isn't great either.
I began a new, long winded method recently. Using the flattest things EVER, the stunt tracks, I created straight test tracks in the sky, as long as the map, and used trainer's speedometer display (the true speed) to reveal speeds. The fact these speeds don't flutter around at maximum revs are already an indication this is the best test method yet. Either way, all the cars survive almost 20 full seconds at a certain 'top' speed before the speed randomly, gradually counts up to a second, slighty higher top speed, at which it will sit at for an eternity.
It's a strange phenomenon. Either way, needless to say, hardly any car's dashboard speed reflect the actual top speed.
I've got a chart going here: http://gta.wikia.com...#Vehicle_Speeds
I haven't begun updating article top speeds. I may begin removing current article speeds soon to suffice the changes to the test and avoid future confusion.