All the big stuff has been summed up, but here's a few details IMHO:
-M1 Garand is about the ideal no-loss platform since the gas port is only inches from the muzzle. Incidentally, as Garand owners know, running ammo with really slow burning powder can cause a bent op rod in that particular design, since it was intended for cartridges with moderate speed powder, that has relatively little residual pressure at the muzzle.
-I would think a 16" AR-15 with a carbine length gas system would be the opposite extreme, since the gas port is so close to the breech that the gas is at really high pressure, and so far from the muzzle that there's some real bullet flight time between passing the gas port and exiting the barrel.
-Old tests on the Garand showed the bullet was something like 20 feet out of the muzzle before any action parts were moving measurably. I expect this is true on most designs, with the exception of AR's having short gas systems and relatively long barrels (again, that 16" barrel with a carbine gas system - a combination that the designer never intended). But even for those I'm sure the bullet is fully out of the barrel before anything unlocks - otherwise cases would kaboom.
-Though I don't have any data, I will bet that simple blowback designs actually show the greatest velocity loss vs. a locked breech, since the case will start moving rearward instantly and the bore diameter x say a half inch of movement creates a much larger increase in volume for gas to move into than a tiny 0.05" gas port into a .25" gas piston or 0.15" (guess) AR gas tube.
-Don't forget that recoil operated firearms can also see a loss, since the barrel will have started moving backward to some degree by the time the bullet leaves. I'll bet the loss is tiny for most of them, but I'd be interested in any actual data. The only easy way I see to get that data would be if you have one of those special ops type pistols that will let you lock the slide in place for use with a suppressor.