It is one reason. But e.g. in my field (theoretical physics) I am not afraid of theft (well, if someone is smart enough to understand and polish a half-written

It is one reason. But e.g. in my field (theoretical physics) I am not afraid of theft (well, if someone is smart enough to understand and polish a half-written paper, then (s)he can write much better stuff) but of things that "as of now makes sense only to me" or "there are half-written" (e.g. in some places I changed indices i to i+1, but not in all, making it both frustrating and useless to read) or may simply be false and I don't want to publish it before checking (e.g. plots may be completely wrong because of a typo). Anyway, a time limit is a good idea. And anyway, when there are timestamps, it's harder to steal an idea. It actually may encourage people to write notes on things to do, share some partial proofs, etc