There are a few assumptions and assertions that probably ought to be cleared up before continuing the wild speculations. I apologize if I get any of this wrong. Hopefully someone on an actual PC can check Wikipedia and fact check me where needed. This iPad isn't very conducive to citing sources on a forum.
Firstly, electroplating doesn't rely on heat or melting metal, but on using electricity and a chemical catalyst to deposit metal onto an object. So heat doesn't easily undo electroplating. That is to say, the heat required to melt rhodium is probably higher than a mod can produce.
I think the melting point of rhodium is, what, 1800 degrees F? I think if an atty arcs against the ELA's rhodium-plated 510 pin, the atty's steel or nickel or whatever will show blemishes long before the ELA.
I know David has talked about all this before, but I don't remember where. I hope he sees this conversation and chimes in. We could use an expert.
Anyway, metals are malleable. They bend, rather than crack. So firing an ELA for 45 minutes wouldn't alter the rhodium and make it brittle or micro-fracture.
But even if the connector is pitted all to hell and back, assuming the atty isn't tarnished or coated in
juice sludge, there's still going to be more surface area in contact with the atty than the diameter of your res wire, so there shouldn't be any current lost there.
I think whatever pitting and discoloration has been seen is probably the same sort of carbon deposit you see on your stove when your spaghetti boils over.
juice gets into the 510 connector, gets in the way of the electrical path, and gets cooked.
But how do you clean that crap off a stove? Since carbon is inert, the only way is to buff it. I just looked at a bottle of Easy-Off range top cleaner. It appears to be a micro-abrasive polish.
But I wouldn't want to make a habit of buffing the rhodium-plated contacts clean. I bet, though, that despite whatever visible imperfections you see, if you attached a 510 volt meter to the ELA and fired it with a fresh charged battery, you'd still see around 4.15 - 4.17v.
I think the only way to resolve this conclusively is with a screw-on volt meter rather than relying on empirical evidence.