You may very well have to slide some tracks.
Hi rharv.
I don't usually disagree with you, but there are just a couple of points...
If we assume Zoom A is the master clock and Zoom B is the slave, then Zoom B is going to run 2ms behind Zoom A. If all the tracks are recorded from scratch with this set-up the tracks will still play back in sync. The only time a problem would arise is if the tracks were transferred in bulk to another medium where they were all running on the same machine. Then the tracks from Zoom B would have the 2ms error. However, we're not told that this is,in fact, the case.
Second point is - can we hear a 2ms error? If you get a guitarist to play a series of notes to a click track and then look at the start time of each note, you'll see variations. Multiply this by sixteen tracks, where some start times are early and some late and it's difficult to spot 2ms overall. This is why we got away with it using tape. In fact, if you record a snare drum hit simultaneously onto tracks one and twenty four of a two inch multi-track tape, you will get time smearing when you play it back.
I will concede that if all tracks are quantized VSTi the situation is not quite the same and I respect your maths for performing the correction.
ROG.