I don’t have an overview of the whole literature but the picture I get is that several scholars quite to very infuential in the respective linguistics here see Pāli as not an actual spoken dialect, but rather as an artificial language, created for example for easier comprehension among the increasingly dispersed communities or the general influence of Sankrit. To mind come:
- K.R. Norman
- O. von Hinüber (for example here: https://obo.genaud.net/moz/resources/pdf/hinuber/vol10_art10_hinuber.pdf)
- B. Levman (for example here: (PDF) The language of early Buddhism | Bryan Levman - Academia.edu)
Don’t really know yet how to assess all these propositions but some observations seem to speak rather against them. For example that Sanskrit might not have had such an influence as to find any access into the Pāli as such.
Mettā