Why two hours of tape is the least useful form available

Duration is the enemy, and an unbroken duplicate of the session is the worst possible default.

Consider what a person is actually doing when they open a long recording. They almost never intend to watch the whole thing in order at the original pace. They are looking for something: the bit about the assessment, the worked example, the part they missed because they were late, the explanation they did not follow the first time. The recording, in its raw state, offers them a single undifferentiated block and a progress bar. Finding anything means scrubbing, guessing, overshooting and starting again. The experience is closer to searching a filing cabinet in the dark than to studying.

So the recording is at its least useful precisely in the form it is most commonly left in. Every minute of unmarked duration is a minute the viewer has to search past. This is why people who work with recorded material seriously tend to spend their effort on structure rather than on production values. Better audio is nice. Knowing where the thing you want begins is transformative, and it is the cheaper of the two by a wide margin.

There is also a quiet cost to the person who taught. A long unstructured file invites the assumption that the session was two hours of undifferentiated talk, which is rarely true and does the teaching no favours. Marking the shape of a session, the point where the topic turns, the twenty minutes that carried the load, is partly an act of navigation and partly an act of honest description. It tells a later viewer what the session was, not merely how long it lasted.