I have used AutolevellerAE quite extensively over the last few weeks with great results.
I have had no recurrence of the issue where there were more data points than probe points.
The problem was discovered and extensively investigated by an ESS user. The problem he described
was not related to probing however the fix for that problem has also fixed the apparently random number
of extra data points. The ESS plugin I'm using currently is 220. The latest release is up to 232.
Given the way in which Mach4 gathers data from the ESS there is the potential for the measured data
points to be slightly in variance with the intended probe points. The macro I have written preforms that
correction nicely.
In the last few weeks the only time I've had any issues was occasionally when a probe point was missing
because of dust/dirt on the board. My macro caught that issue and reported the error.

Interestingly enough the main Mach4 developer is making progress with a 'surface probe, mapping
and correction' feature coming up. It will be interesting to compare approaches.
Craig