I like to help with doing IDs on iNat and I agree about the two IDs. I think that is a potential issue for trusting the data because I know sometimes I see research grade observations that are incorrect. If I’m using Identify mode…which for me is the fastest way to knock out a lot of IDs for iNat…then I would never see the incorrect RG observation because Identify mode skips Research Grade obs. This means that “bad” data runs the risk of never being seen or found and gives credence, perhaps, to experts not to trust iNat data.
One example of this was Chilean Nailwort (Paronychia franciscana) that many people misidentified as Sandcarpet (Cardionema ramosissimum)…which resulted in Sandcarpet coming up as the first suggestion for the nailwort (and to make it worse, since the nailwort did not have enough IDs, it never even showed up in the list of suggestions at all…so the incorrect RGs kept happening). I finally spent a lot of time trying to go through all of the Sandcarpet IDs in the location where the nailwort grows to correct them (or at least give another ID to bump them out of RG in hopes iNat would stop suggesting Sandcarpet for nailwort). I am seeing less mis-IDs of nailwort now so hoping it worked.
I will add that I am NOT an expert so I try to stick to only plants I know very well, but there is a chance (very likely just for the fact that I’m human) that I will make mistakes at times. The “RG after two" IDs” makes me often NOT ID things that I’m pretty sure I know because I don’t want to bump it to research grade incorrectly. Maybe if I knew it would get one more pair of eyes on it (other than the owner of the obs), I would be a little more brave (this could be good or bad, I suppose 
).