Thanks.
The Tri-tag article mentions beacons for the ATP system, but my understanding was that ours leverages the AC signals in the crossing prediction blocks. There certainly aren't any conspicuous fobs bolted between the rails like you see on Line 1, Line 5, or in Ottawa. I believe that's what it made it particularly finicky to tune initially, and why sections like southbound approaching Hayward got awkwardly slow once it was switched on, and have never been fixed.
If that's true, then I suspect there's no way to fix it without splitting some of the signaling blocks, a major infrastructure change that may not be feasible without renegotiation of the Design-Build-Finance-Operate-Maintain (DBFOM) contract, and would also require a (lengthy?) bustitution period.
I sincerely hope that my understanding is wrong.
At the very least, I wish they would tune the crossing predictors on Courtland and a few other notorious spots on the line. The gates go down far too early now and stay down far too long, but they weren't that bad when the system opened. The analog parameters of the track have changed as it's built up grime and worn in, but it seems like no one's ever gone back to tweak the programming to align with reality. Oh happy day when we finally see the clipboard and stopwatch crews out there again like they were during commissioning. (Although you'd think that just videoing the operations in various weather conditions and frame counting would be sufficient -- but I have no idea what the rules 'written in blood' actually are.)