If the plate scale was way off that would be the probably reason for the solve failures. The tiny adjustments between the uncorrected and corrected guide star positions shouldn't make or break solving. For background:

The confusion stems from the old days. Bill Gray produced a re-mastered GSC/1.1 in which he applied the corrections for systematic errors he found in the original catalog. This "GSC-ACT" catalog was significantly better for accuracy. So in a program designed for the straight GSC/1.1, you could use the Bill Gray GSC-ACT in its place and enjoy that increased accuracy.

I got the corrections from Bill and put them into PinPoint as run-time corrections to be applied to the vanilla GSC/1.1. Since PinPoint was designed for science work I needed to provide a way to solve with the uncorrected GSC/1.1 as well (traceability, etc.). Therefore I put in two different catalog modes for the GSC/1.1 catalog. In the API, I labeled the one for which the Bill Gray corrections are applied as "GSCACT", and the other as "GuideStar" (see below).

I'll suggest to Steve that he simply provide "Guide Star Catalog 1.1" as a choice and set the Plate.Catalog to ppGSCACT.