Thank you Brian and Roland.
While its not impossible that DEC balance was off for some reason, I would say it's unlikely as I usually do it the same manner Roland described and last night was no different. In fact, I even use the dovetail riser as a reference point (see below) when loading the OTA which I know is pretty spot-on balance. But I always check in Park2, before balancing RA, swing it to Park3, balance RA and then send to Park5 to check again and adjust if necessary.
Unfortunately, as I setup and tear down every night, I can't assess if that was the culprit. Will need to wait for the next clear night to see if any different, hopefully by the end of the week.
Here's a shot with all 4 DEC clutches fully loosened and OTA at balance point above. I can easily move it up and down with a single finger by lightly pushing on the top/bottom side of the dewshield or backend of the camera and it won't freefall in any case.
I also should say that I did a Guide Assistant run at the end of the night like you suggested, and PHD2 then reported a 20 arc-min from NCP !
Now, I'm not clear on how such a difference (from the beginning of the night, where it supposedly was in the 2-3" range) could have occured as I almost overtighten and triple-check each and every knobs, clutches or whatever I can grab onto before firing up the imaging session. I make sure of that because temperature sometimes falls quite suddenly at night at this time of year (20C in like 1-2 hrs) and I've experience times where metal contraction was enough to virtually loosen things up. Last night was not that bad (about 10C between evening and midnight) but maybe I overlooked one of the dovetail knobs or the wheels that tighten the altitude adjustment, but again I'd be surprised.
Another thought I had today, was the possibility that the Polar Drift Align routine in PHD2 might have had difficulties in evaluating the field rotation correctly. Especially since the field of view through the OAG is tiny at 1422mm (like under half a degree with QHY5-174MM) in addition to the serious wind gusts that were relentlessly pushing against the assembly and then, encoders fighting back to keep it on target as you saw on the erratic initial Guiding Assistant plot.
In the end, could it be possible it might have mistakenly directed me away from the pole instead of towards it ?
The thing I am uncertain about that theory however is how that would explain the 130s periodic spikes I got for a while. If my PHD2 settings were reasonable and Dec balance OK, then I just hope I have not prematurely worn out the gears/motors/else by accident in some way. I'm usually pretty careful, but life taught me that too much care can end up in the entire opposite of what you're trying to achieve sometimes... :-S