Hmmm... I am not aware of ProteinPilot being able to deal with SWATH data. If you were using a SCIEX application for SWATH data I would expect it to be SWATH 2.0 in PeakView. Between Skyline and PeakView I would expect results similar to those detailed in the Navarro, Nature Methods 2016 paper (https://www.ncbi.nlm.nih.gov/pubmed/27701404), which I will summarize as very similar with a noticeably smaller number (10-40%) of detections between PeakView and Skyline/Spectronaut/OpenSWATH (which were all very similar).
If you are actually trying to run ProteinPilot on SWATH data, then it may not be able to handle that. At least I am not expert enough in ProteinPilot to give you more information on how to use it correctly for SWATH data, but if you used it directly, say as you might try to use SEQUEST or Mascot on SWATH data, then you should expect that it would find very little because of the high degree of chimaricy in SWATH spectra (i.e. most spectra are composed of multiply isolated precursor ions).
Hope this helps. You might also want to seek support for SCIEX on this.