But even the most expensive mounts/piers never will be completely rigid and in addition a small imbalance is desired to improve the guiding, so that there is always a small force against the leading or trailing edge of the worm.
As long as worm gear mounts are used without spring pressure on the worm, probably the backlash as result of using symmetrical positions before & after the meridian, will make a greater contribution to the PAA inaccuracy than the flexing caused by the imbalance. And of course FL and the weight of the telescope will play a major role.
However, with spring loaded mounts there seems to be an advantage by using symmetrical positions before & after the meridian to polar align, no matter how small the flexing / bending may be.
First, RA backlash is irrelevant. There is no requirement for accurate movement in RA, you just need any three positions. The selected step size, AFAIK, isn't used anywhere in the calculations.
DEC is another thing, so if there is substantial backlash that might affect the result. However, the assumption for the algorithm is that there is no movement in DEC, so in that case the asymmetric version (keeping the DEC gear always on the same edge) would be the better choice.
Imbalance of the scope will only affect gear meshing, on which edge you sit. I don't buy that even a more severely imbalanced scope will cause a noticeable static bending.
Or, the other way around: If it's so imbalanced that it causes bending, PA will be your smallest problem....
That said, I'm always doing my PA symmetric, starting at -45⁰ going to +45⁰
And I don't move back to home/park afterwards, and only revert direction for the next run (if needed).
openSUSE Tumbleweed KStars git INDI git
GPDX+EQMOD, CEM60EC, ASI2600/1600/290mini+EFW+EAF