Refactor `convert.py` with greater modularity

> I agree with @gwaybio's comment that `convert.py` is getting long and it's a little difficult to focus on the parts that matter, but since this PR touches very little of that code I'd refrain from doing that refactor here.
>
> For when you do refactor it, I'd prefer to see the utility functions moved out to different modules, but I'm unsure if that would complicate things with how parsl ships out modules to its workers. You might be able to get away with reordering things in the convert module, i.e. moving `convert()` up nearer to the top of the file; Python doesn't require that all functions called within a function are defined when it's defined, just that they're available when the function is called.

_Originally posted by @falquaddoomi with references to comments by @gwaybio in https://github.com/cytomining/CytoTable/pull/424#pullrequestreview-4010359644_

Start by reading convert.py and the linked pull request discussion, especially the concern about how Parsl ships modules to workers. Identify the utility functions and check how convert() depends on them before choosing a modular layout. Done means the refactor improves focus without complicating worker execution.

Written by the indexing model from the issue text.