I've just done a flawless TLI with that scenario. Very close to free return and MCC-2 would have about the same DV as the flight plan.
Your pre MCC-1 scenario presents a bunch of clues what could have happened, in the LVDC variables. My conclusion is that up to 7-8 seconds before cutoff everything went normal. The LVDC went into terminal guidance at 15 seconds before cutoff. But it never went into the attitude hold mode for the last 3 seconds of the burn. The last time-to-go until cutoff the LVDC had is 7 seconds. The cutoff was not caused by the LVDC with a velocity cutoff (nominal method) and it also wasn't caused by the S-II/S-IVB staging switch or a cutof signal from the CMC.
So the only things left I could imagine is a manual cutoff with the EDS or the Orbiter method. The THC can be used to shut off the engine like an abort. And what sadly hasn't been disabled yet is the plus and minus keys on the numpad. They work like for all other vehicles in Orbiter, + gives you full thrust, - cuts off the engine. I've tested it and I can generate about the same conditions in the LVDC as in your pre MCC-1 scenario by using the minus key a few seconds before cutoff. So maybe you accidentally hit the minus key?
In either case, I think you should just try that pre TLI scenario again. If everything works fine (EMS went to -80 for me, that is a good sign) then you can just continue. If the same issue happens again, maybe the LVDC log has some clues. So after TLI fails again (if it fails) please post the file "lvlog.txt" here so that I can take a look. That file in the main Orbiter folder.