refactor(web): extract the correct-and-raw-predict blocks into their own method 📚

[jahorton]

Continuing from #11899 (spun off from this one), this PR makes the correct-and-predict section its own method. Even on its own, the new method is pretty large, but at least its bounds are made distinct from the rest of the sections by doing so.

We can also shift to a simpler + clearer pattern for 12.0 models vs 14.0 models by returning early when the more complicated process isn't viable, which I believe does result better clarity here.

Note: reviewing with "ignore whitespace" appears to make the reviewing process much simpler for this PR.

@keymanapp-test-bot skip

[keymanapp-test-bot[bot]]

User Test Results

Test specification and instructions

User tests are not required

[keyman-server]

Changes in this pull request will be available for download in Keyman version 18.0.75-alpha