Thanks for the explanation! Yeah, I guess it's a similar or at least analogous situation for Proto-Indo-European, where there might be conventional ways to pronounce reconstructed forms, and some evidence about what is more and less likely, but people don't really know. There are even some reconstructed consonants where they actively have no definite consensus at all:

https://en.wikipedia.org/wiki/Laryngeal_theory#Pronunciation

I guess it's quite an imperfect analogy, because we have Akkadian writing (by people who actually spoke the language) and we have no equivalent at all for PIE.