OK, as much as I really thought (or hoped) I would have no reason to revive this thread, here's the thing...
I've been using the above referenced Xstream on my left post when diving doubles (so far without further problems than the ones already solved in thread), while using an Apeks MTX-RC on my right post. I decided to get a second Xstream when the opportunity came up in order to have a matching set. It is pretty much an identical unit to the one I originally purchased, a brand new unused Xstream black, that I bought from someone on FB who had purchased it for travel a year ago, but due to COVID, was not able to use it and decided to sell. Anyway, the regulator has never seen water, and was manufactured in September 2019 based on the serial. I proceeded to put this reg on a tank to test the IP as soon as I received it, and guess what: it has almost the exact same problem as the first one. On a full tank, the IP initially settled in the 130-140 range, and then very slowly kept creeping up to settle a little above 150 psi. Pretty annoyed at this finding, I implemented the strategy outlined above by
@rsingler to see whether dust could be the culprit and be in the way of the HP seat sealing correctly. I opened up a low pressure port and cycled the tank valve on and off a bunch of times to attempt to dislodge any fiber or other debris that could potentially be inside the chamber. When I retested the regulator again, the IP was now bouncing to ~125 after a purge or breath, but kept creeping very slowly (one hour plus) to ~ 140psi before finally settling. I thought the problem was getting better, but when re-testing on a full tank, the numbers are about the same as the initial test. So it seems that the initial value to which the IP rebounds and how much it creeps overall depend on tank pressure, with a final resting value of ~150 or so on a 3300 psi tank, and maybe 130-135 on a 1500 psi tank or so.
What is going on here? This is getting kind of ridiculous. Either my garage is like the Bermuda triangle of Poseidon regulators or there is something fishy about Swedish quality control...
By the way, my gauge works just fine. My single tank setup leg3nd locks up rock solid at 122-23 psi, so my gauge is certainly not over-reading.