Thanks for the detailed replies. For my case, I do have excessive backlash or rather, loose gear coupling, since I purposely left the RA mesh extra loose with an East heavy weight bias to see if I could reduce the magnitude of the RA error spikes. For that test, I measured no guiding improvement and actually saw the mount wiggle with the breeze, since I have an 8" Newtonian telescope masquerading as a sail. In the end, I'm not seeking perfection and complete reduction in backlash, however, I am looking for a robust solution that allows me complete remote control and functions well on both sides of a meridian flip without manual intervention. Also, I've got a total payload weight of 27 pounds on the EXOS-2GT, so I'm pushing the mount beyond what it is really designed for. That's partly why I've completely tore down, tuned & rebuilt my EXOS-2GT mount and am systematically working my way through every part of the Astrophotography process to identify what significantly impacts guiding performance and where the most improvements can be found.
Regarding polar alignment - I also setup and tear down my equipment every session and do not have a clear view of Polaris in my backyard, so I do a rough alignment with my cellphone, then do a drift alignment with PHD2. I can usually get < 200" of PA error in around 5 minutes, since I'm primarily only refining the azimuth adjustment.
I am curious as to how long you are able to guide with both the EXOS-2GT and the G11?
Mount: ES PMC-8 EXOS2-GT
Scopes: BRESSER 208mm f/3.9 Newt + ES HR Coma Corrector, Celestron C90 Mak, generic 50/191mm guide scope
Cameras: Canon 5DmkII, Canon EOSm, T7C (ZWO ASI120MC clone)
Misc: 2" Optolong UHC filter, Raspberry PI 4
Software: INDI + Kstars + EKOS, DeepSkyStacker, GIMP, Lightroom