I notice that when I mouse over to and click on the autopilot pitch controller on the radio popup window, that the first click in the opposite direction is always in the wrong direction. For example, if my last adjustment was to increase the pitch to, say 10.0 deg, if I move the mouse off the controller, then come back to it and click on the pitch DOWN side, the first click results in the pitch target going from 10.0 to 10.2 deg up...subsequent clicks then start to decrement the value. The opposite also holds true...if last adjustment was down, the first click on the up side will always be another click towards pitch down even if I'm clicking on the pitch-up side.
Also, placing the yaw damper switch guards down does not force the yaw damper switches on...it should not be possible to have the guards down with the switches in the off position.
Second issue is the very erratic TAT and IAS indications during climbs and descents. MSFS has a known problem where the OAT is only updated once per second. Each time the OAT updates, it changes IAS, mach and TAT...during climbs and descents, when the OAT is steadily changing, this once-per-second update rate induces a nasty tick in the gauge display. FSUIPC has a smoothing feature built into it that causes a moving average to be displayed...this effectively dampens out these erratic "ticks" in the IAS display. If the panel already uses FSUIPC, it'd be a simple matter of having the gauge get IAS from FSUIPC rather than directly from an FS callback...if not, I think it'd be worthwhile to try and smooth the indications within the gauge code in much the same manner as FSUIPC. I believe the details are in the FSUIPC SDK.
ATP IMEL Gulfstream II-III-IV-V L-300