Still running a little "off" after installing OEM coils - ignition timing jumping around at idle - ideas?

I finally got around to calling up Quick Tyme and getting some OEM coils for my '13 turbo.

The car responded as it usually does to a coil change; get glassy smooth to start and then have roughness slowly set back in afterwards. I will say that the roughness isn't nearly as bad as it was on Duralast coils, however, it still does vibrate a little at idle and at certain RPMs, and sometimes when idling it'll have momentary RPM drops or even try to chug out.

At idle, I usually see the spark advance bounce around +/- 3 degrees from 0.0, with occasional trips outside of that - I want to say that a smaller range correlates with a smoother idle at the moment, but I haven't fully confirmed this. Well, this time I finally got my OBDII scanner plugged in and mounted while it was doing the rev drops. It looked like the spark advance was soaring to like 16 or 17 degrees for a brief moment before cratering to negative double digits, like -15 or -20 (I don't remember the exact number on that one), then the numbers and the idle would both normalize.

These rev drops only happened once today, immediately after filling up 90 octane at a Fred Meyer gas station. Otherwise the car was very smooth, though still with a bit of idle roughness.

Possibly related: sometimes when upshifting very fast (but not flatshifting), there will be a palpable - and inconsistent in terms of duration - lag upon completing the upshift before the engine will deliver full power again. This also needs more testing to suss out (it's hard to watch a scanner safely while doing a WOT pull), but it appears that the spark advance may be taking a brief moment to reach its usual WOT value, which IIRC is either +9.0 or +9.5 based on my readings). Sometimes I'll get other forms of inconsistent power delivery, but have never been able to correlate these with spark advance.

I'm going to try, sometime this week, to re-gap my plugs to 0.028 (currently whatever they came out of the box at, which as far as I can tell is 0.031, and they are the correct NGK Laser Iridium type), but until then, does anyone know what might be causing this odd behavior?

For reference I'm at roughly 27K on a new short block + refreshed top end, and consuming no or nearly no oil.