feat: proper implementation of FLEx pronunciation field that supports audio S/R

[megahirt]

Is your feature request related to a problem? Please describe.
LF does not properly interop with the FLEX pronunciation field

Describe the solution you'd like
This is a known issue already documented in our Known Issues Wiki, however there wasn't a stub issue created specifically for this, so I am creating it now.

The problem is that the LF model for pronunciation does not match the model that FLEx has implemented, resulting in a non-functional pronunciation field in LF.

Pronunciation in LF needs to be a "list" model, similar to senses and example fields.