The staff time that nobody puts in the budget

Everything worth doing to a recording is labour, and it is currently being absorbed rather than funded.

Adding chapters takes time. Correcting a transcript takes time. Writing a short summary takes time. Checking whether last year's file still describes the current requirements takes time. Deciding whether a recording of a discussion should be released at all takes judgement and therefore time. Individually each is small, which is exactly why each gets assumed rather than allocated, and collectively across a teaching load they are not small at all.

The predictable outcome is that the work happens where somebody is conscientious enough to do it unpaid, and does not happen anywhere else. Quality across a collection becomes a map of individual goodwill rather than a consequence of any decision, and the people carrying the load are usually those with the least secure arrangements and the least capacity to decline. That is an unstable arrangement and it does not survive a busy term.

Naming the work is the beginning of fixing it. Decide which recordings warrant the treatment, because not all of them do and pretending otherwise guarantees that none get it properly. Decide who does it and count it. Accept that a smaller number of well structured, well described, current recordings is worth considerably more than a complete archive of raw files that nobody can navigate. The choice is not between doing this and not doing it. It is between doing it deliberately and having it done invisibly by whoever cannot bring themselves to leave it undone.