I begin this post by apologizing for not completely reviewing the thread again before commenting. I did read it while the incident unfolded in the months afterward, but the thread is now so long it precludes the ability to find specific information for memory refreshment.
There has been discussion about the controller's workload; at the time this individual was coordinating aircraft flight and ground operations as well as the movement of ground vehicles.
My thoughts on the issue:
The aircraft was cleared to land and on a visual approach. It therefore seems to me the controller's workload had little bearing on the crew's lineup on the taxiway rather than the runway. Similarly, the controller's go around call was based on his interpretation of a radio transmission from the crew of UAL 1, not information that was available from his equipment.
It seems quite reasonable to state the controller's increased workload had little effect on the situation.
As for the crew's perception of the last seconds of the approach, the situation was undoubtedly confusing.
"The NTSB reported that in post flight interviews both pilots of AC-759 reported they were convinced the lighted runway to their left was runway 28L and they were lined up with runway 28R. They did not recall seeing aircraft on taxiway C, however, something did not look right to them."
Runway 28L had been closed for five months at the time of the incident. It seems reasonable to assume the crew might have flown into SFO during those months, and therefore on the date of the incident might have been aware 28L had been closed for some time. The NOTAM regarding the runway closure was part of the data the crew reviewed that day, and the ATIS broadcast also contained the information.
In spite of this, it seems the crew did not fully integrate this knowledge into their approach briefing and actions during the approach. The NTSB statement above says the crew mistook the lights of Runway 28R for the closed Runway 28L. But as they should have expected, the closed runway did not have approach lights, PAPI, edge lights, and centerline lights in operation. It did have an elevated brightly lit "X" placed on the centerline of the runway. Why would they make this elementary mistake?
The crew was aware the 28L was closed, and the difference in the sight picture between a fully lit active runway and a taxiway, illustrated by an earlier post of a sim program screencap, is so obvious it seems inexplicable the crew made the misidentification described in the NTSB document.
The inquiry the crew made to the controller about their confused sight picture seconds before the planned touchdown indicates they were aware something was amiss. But because of their alignment on the taxiway, the ASDE-X/ASSC data link did not show the position of AC-759 for several seconds, leaving the controller unaware of the impending collision.
The assertion by both pilots they were "convinced" the aircraft was aligned with Runway 28R seems to be a bit too emphatic to me. If so, why were they confused enough to contact the controller about the situation? "Convinced" conveys the image that they had a clear understanding of what was in front of them, and that characterization seems to be, perhaps, rehearsed.
While giving sufficient weight to human factors, particularly fatigue, the actions of the crew seem to indicate inattention or complacency were also present. This should not be a surprise, human endeavours are often affected by such traits. I have not made the above observations to heap unwarranted criticism on the crew, I am suggesting behavior, though unintentional, that may have played a part in the incident.