Lewis & Short cites a few marginal ante-classical uses of misereor with the dative. Perhaps this was simply a more colloquial or rustic usage that survived and reëmerged in the Ecclesiastical era. I don't know.
I think your original interpretation of the difference between quot and quantus,-a,-um is the correct one, and I'm not sure how to account for the fact that M&F uses quantis diebus in their example. Just by doing a Google search for quantis diebus and quot diebus, it becomes obvious that the latter is far and away the more standard usage, whereas the former only appears in a few non-classical sources, including M&F itself.
Lewis & Short has only a very short section citing uses of quantus,-a,-um instead of quot for countable quantities. The few it does have are mostly poetic, e.g. quanta milia for "how many thousands" in Propertius. These may have been influenced in part by Greek usage, as Greek uses ὅσος for both "how much" and "how many".
The example of quantas pecunias in Caesar appears to mean "however great sums of money" rather than "however many sums of money". Of course, the neuter quantum may be used by itself to mean "how many" in addition to "how much", since a partitive genitive can always be implied, e.g. omnes di deaeque, quantum est [eorum]... "all the gods and goddesses, however much [of them] there is...". But this is quite different from what we see in your example from M&F.
Ex mala malo
bono malo uesci
quam ex bona malo
malo malo malo.