P(infect another) = P(get the disease)*P(transmit after getting it)
With the Ebola example, that's P_Ebola = 1.0*P(transmit after getting it).
With the COVID example, that's P_COVID = 0.3*P(transmit after getting it).
Both of these are sub-1 and not guaranteed to happen. In fact, P_COVID may actually be larger than P_Ebola depending on the relative R0 and assumptions we plug into the above calculations. So I don't see why the distinction you raised has relevance.
Detaining someone with Ebola, which we both agree is a good idea, is also a pre-crime measure, because they haven't infected anyone yet and they may never have infected anyone even without detention.
And I will also point out once more (due to your second sentence) that I do not support forced vaccinations for COVID. The COVID disease just isn't deadly enough and the vaccines aren't effective enough at preventing transmission for me to support that. I am merely saying I do support it in the abstract if a sufficiently deadly contagious disease comes about in the future.