Antenna calibration... or clock calibration indeed?

Hi all!
When I calibrate antenna delay, do I really measure radio path penalty we all have to consider, or particular anchor’s clocks vision of this problem? What I really care about, is the extent, when error in calibration of particular device is below of errors, forced be clocks drifts. Indeed, I observe not that fast mutual drift of clocks in my swarm of similar devices, so does it realy worth to sqeeze 10-20 ticks from calibration of particular antenna?