Just use Crystaline panels and the problem goes away.
It's more nuanced than that. There were a lot of PID study results presented as part of Solar Power International's "poster program" in Chicago. It's fair to say that there isn't any consensus RE: root causes of PID. But it is a real concern, and it definitely impacts c-Si PV modules.
Based on what I've read, PID impacts in the field will vary based on: module construction, system grounding, array voltage, module mounting, relative humidity (high humidity environments are worst), etc.
A colleague of mine noticed language in SunEdison's warranty restrictions that seemed to rule out using their modules in bi-polar arrays. When he contacted the manufacturer, an applications engineer confirmed that the power warranty would not apply if these particular modules were used in a bi-polar configuration. My impression is that PID was the reason for the warranty exclusion.
This white paper by AE is relevant:
http://solarenergy.advanced-energy.com/upload/File/White_Papers/ENG-PID-270-01 web.pdf
The AE white paper seems to suggest that PID effects are module specific. If so, you should be able to find a c-Si PV module manufacturer that is more than happy to give you a thumbs up to use their product in a bipolar application. I'd go that extra step before placing a PO.