Well, tbh... when your measurements are really correct it should be ok. Problem is that in most situations you cannot really measure to the voice coil.
F.e. in my Mustang i have to estimate the distance from the grills to the voicecoil of the tweeters and the midbass drivers.
And i measured the distances wrong on several occasions. When you measure correctly and your estimation of the distance from grill to vc is almost spot on there should be almost no difference to the IR measurement.
And tbh... i doubt that a lot of people can hear the tiny differences at all, even when you're missing the measurement by about an inch or so (which is a lot for tweeters due to short wavelengths). On the other hand... do these differences really matter when just moving your head from left to right totally fucks up your phase from your tweeters response?
I'm just doing measurements for the IR, because i like the concept and tinkering with it and to get it as perfect as possible with my very and totally limited knowledge that i gained so far.
So all in all Andy Wehmeyer is for sure right with what he's saying. IRs, when measured correctly will always show the same delay between speakers when all speakers are set to 0ms delay. Same is true for correct distance measurements. 50inches (or the corresponding value in centimeters where i live) are always 50inches if you don't move the driver to another location. It is 50 inches now and will be 50 inches in a week. So yeah. Measure carefully, type it in your DSP and forget about this. Especially if you can alter the phase in parallel to the distance like you can with the Helix processors.