That is just the point though. While it is possible that something might have changed, it is also possible nothing has changed. The difference between 1000 and 1500 idle RPM when AFR, timing, air temp, battery voltage, age of oil, fuel variations, CLT, MAT, underhood temp, ... is not much. I would think that 1 point AFR, 5 degrees of timing, and 5kpa MAP would be more than sufficient. Be aware that at MBT you will require a lower MAP to maintain a higher idle speed.

You cannot assume that the operating conditions are the same between now and then. If you have similar data showing no change from before you should look at that.

Ignition timing is the most important parameter.

This type of testing should always be conducted at full operating temperature.

At one time I posted characterization data showing idle RPM as a function of spark timing over a range of either AFR or manifold pressure, possibly even both. It is here on MT, not sure where.

Ok, fair enough. I still feel like something whacky is going on to cause that blip, but I have checked a bunch of other things and am still seeing the RPM oscillation so will now attempt to tune closed loop idle.

Things I've checked over the course of this thread:

1) Checked for vacuum leaks by spraying brake cleaner around all known vacuum lines and junctions and looking for pickup in engine sound. None found, though maybe I missed a spot?
2) Checked for grounding issues caused by the throw-out bearing making contact and closing a poorly-grounded/ungrounded circuit
3) Unplugged the neutral safety switch to make sure it was not passing along voltage to the ECU (IIUC service manual says that it sends 1V to ECU if in gear, 0V if in neutral). Just for kicks, I also jumped the connectors to simulate "fail open" and tested again to see if it made any difference. No change to the RPM oscillation, so unless I somehow disco'd the wrong thing, the NSS is indeed not the problem.