PIE * p has widely become h in Armenian (e.g. հարց (harts) "question" < * prsk-, հուր (hur) "fire" < * pur-, etc.). However, some have claimed that the verb փլիլ (pHlil) "to fall in, collapse", which begins with aspirated or "fortis" p, comes from IE * pHol- "fall", i.e. a root with aspirated p.
Regardless of whether this etymology is correct (I know that there is some debate around whether PIE had voiceless aspirated stops to begin with), I would have thought that aspirated [p] would be at least as vulnerable as unaspirated [p] (if not more so) to becoming [f]/[ɸ] and subsequently [h]. But, even if there was no intermediate stage of [f]/[ɸ] in this case (I'm not sure how probable p > h would be without such a stage), is there any reason to think that aspiration would make [p] more rather than less resistant to the process that leads to [p] becoming [h]?