Well, taking APCC out will not necessarily eliminate APCC, the AP driver, or SGPro, if the problem doesn't happen without APCC, because the interactive timings will have changed, causing the scenario to change. It doesn't necessary identify which component is at fault in the original case. Only if the problem still happens can you 100% eliminate anything (APCC in that case), but it could then it would leave the driver and SGP.
That said, there are at least two APCC problems here. The first I think needs to be solved within APCC:
1) APCC's UI thread appeared to be stalled for a period of time. The obvious solution is to move virtual port return processing into a background thread. It was just very convenient for it to be in a UI thread because it makes it very simple to update UI controls. But doing something in a programmatically simple way doesn't always produce the best user experience.
2) Why did SGP think that the slew was done early? (is there a bug in the AP driver, or was it because of the APCC/driver "UI stall"?)
-Ray