I talked to Jim Finch at Tascam in CA.
Regarding the tape tension issue: He suggested I check to see if the tension arm spring on the supply side is stretched, or even to try putting a second spring on it (there are two connection points on the hanger, BTW). He said it wouldn't be an issue with the reel motor, especially since the connector I added on the motor connection was brand-new. He said that even if the connection was really corroded it wouldn't be enough to cause the kinds of tension issues I am seeing...there would be bigger problems first. He suggested to check the tape to head contact again on a 16kHz tone (i.e. playback 16kHz and lightly apply pressure on the supply flange in PLAY and watch for meter deflection especially on the outside tracks). I did do this initially, and it was fine, but I think I did it on a 10kHz tone and it was prior to all the adjustments that I have made. Either 10kHz or 16kHz will work but he says the 16kHz tone is more sensitive. If, after making all these adjustments, the levels are stable with increased back-tension on the supply side, then there is sufficient tape-to-head contact and the 60g tension on the supply side (as opposed to the specified 75-85g) in PLAY is not a problem. In fast-wind the numbers aren't in spec to be sure in REW, but basically he confirmed that I should be more worried about tension that is too high vs. tension that is too low, which would effect the tape pack, but since I'm spooling to the takeup side and storing tails-out, and my tension readings are pretty good in FWD (and both REW and FWD tape packs look good anyway), that I may indeed be beating a dead horse. That is what I was looking for. So I'll check the tension arm springs (I'll swap 'em and see what happens), double-check the tape-to-head contact, and if there is no change with the spring swap and the tape-to-head contact is good then I'm leaving it be.
Runaway slave issue: it does indeed sound as though, somehow, the ES-50 synchronizer is not reading the tach pulses from the 58. I did confirm some time ago that I'm getting the 12Hz tach pulses at the ES-50 end of the sync cable in PLAY, approximately 72Hz in SPOOL FWD or REW, but then I'm getting garble in fast-wind. That was with my previous DMM. I'm going to retest with the Fluke 85. If I'm not getting a reliable frequency in fast-wind at the ACCESSORY port then there is a problem inside the 58. If I am getting reliable pulses at the ES-50 end of the cable then there is a problem with the ES-50 and I will likely need to send it to Tascam.
The other piece here is that the ES-50's setup/calibration procedure (which auto-sets synchronization parameters specific to the slave and master machines to which it is connected) is not able to complete with my setup because I only have one machine (the 58 as slave) attached. Jim is going to contact engineering and see if there are steps that the ES-50 ignores accross the board if both master and slave machines are not connected. In other words, the auto-setup/calibration procedure on the ES-50 was designed to function with master and slave machines connected. If there are parameters that cannot be read from a machine the ES-50 ignores those and tells you that it could not complete the setup. That doesn't mean it stopped at that point, it just couldn't complete those particular steps. The question is whether or not in my case, where I have a "code-only master" (no master deck is connected, just the SMPTE code from the DAW in the MASTER TC input), none of the master machine parameters will be able to be read, since there is no master machine. So will then, for instance, certain slave machine parameters be ignored because there is no master machine connected? That's what Jim Finch is going to try and find out. If I don't hear back from him it means that he found no more information. So I'll test and see what I can find and then go from there. Ultimately it may mean that I need to refurb my 48 and build the master sync cable before I can properly setup the ES-50 and efficiently get into syncing the 58 with the DAW.