Just want to be sure you are reading the VSWR right. When you say it is up to 2:1 with the splitter in place, it is almost into the "red zone" on the little meter in the software
, right? If it is really 1.2:1 and not 2:1, that is not a problem.
An SWR of 2:1 by itself would not significantly reduce signal levels in and out of the transponder. That level represents only a 10% level of reflected power, and it would be hard or impossible to detect a practical reduction in performance at that level.
If what you are saying is correct though, I would certainly agree there is some kind of problem somewhere between splitter input and transponder input, and the 2:1 SWR is an indication of that problem, though itself not the actual cause of a noticeable reduction in performance.