folks, look at Quiet Bridge procedure, the offset is to threshold. Note the runway is at 284, while the approach is 275 until threshold.(not to 4nm as I have seen posted) As a reminder, it is a visual approach.
Thus, you will be off on the PAPI until threshold if you follow the procedure. That being said, the ac in question offset to a point far out from that, and was aligned with the taxiway for an extended timeframe, even questioning lights on the runway. The ac never deviated from the alignment on the taxiway. (even when GA, and it is not clear if the ac initiated GA before it was instructed to)
What is apparent, at least to me, is that if the taxi had been clear of other ac lights, it is very, very likely the ac would have landed on it. TCH is 55 on 28R, so a GA at 100 (as reported) means that, well, as much as many want to say this was not an issue and the crew recognized and did a GA, well, damn, they were going to land, face it. (when you GA at 28R, you turn to 265 on GA...no deviation in that FP by the crew.
In regards to the visual comment at the beginning, SFO has and does use ADSB. Comments regarding ADSB use, and how it may have saved this are not accurate. Note that ADSB, (out and in) with latency, is about useless on final in this case. Also note, that many ATM systems, in order to remove clutter from the ATC screens, remove the ADSB data (due to clutter and accuracy issues at that point)