Strong agree that those documents had the potential to lead to the murder of informants - this alone convinces me that the mass release of those cables without any edits or redactions was irresponsible.
However, I've previously assumed that we (incredibly) made it through the aftermath of the leaks without any of those potential murders taking place, for the same reasons listed in the grandparent comment; the media would have reported it, and the government would have surely cited it at Chelsea Manning's trial. So I'm skeptical but intrigued by your assertion that informants may have been murdered, but that there's be no way for us to know whether this had happened. My (uninformed) expectation would be that we WOULD know, since:
- we had (previously but not anymore?) informants in those places in the first place
- such murders would be (easily?) dicoverable after the fact, especially since the Taliban would have motive to tout them
- the government had a strong motive for wanting to portray the leaks in the worst possible light, and therefore would want to discover and verify such murders
Your observation that we have very little knowledge of these places does make me more uncertain about whether any murders occurred, but at this point we're far enough out from the leaks that the burden of proof lies with anyone asserting that murders probably DID take place, for the reasons listed above. (As opposed to the question of whether the leaks were irresponsible in the first place, which for me is a resounding yes even if we somehow made it through without any murders.)