Reporting back on the OVP issue to add a data point to the problem. Now a couple weeks since I replaced the OVP, and the problem disappeared and has not reoccured. Before and after the repair I have replaced ALL the rubber in the idle circuit including the bottom plenum of the mass-flow meter on, with varying effects on the idle. The OVP was replaced separately from any other repair in the sequence, and resolving the issue definitely dates from then.
I assume resistive contacts, from whatever cause, were the issue with the old OVP. I tore it apart with additional resolve, and found little inside to indicate a serious problem. Still, I resoldered the pin contacts, removed and disassembled the relay for cleaning (some contact pitting), and will keep it as a spare.