Miscellaneous UI Flow Improvement Suggestions

[HandOfCthulhu]

• Move "Current Speaker" selection to right UI bar to co-locate with (directly above?) the TTS Settings that are associated with them.
• Combine "Current Speaker" and "Manage Speakers" by just adding a + and - buttons next to the Current Speaker dropdown. Speaker/Manage Speakers menubar items can be removed.
• Instead of the "Regenerate Mode" and row coloring conflicting with speaker association, etc. add a "Generate?" checkbox column to sentences table. Sentences that do not have an associated generated audio file are locked as checked. They are unchecked as the sentence is generated and associated. User can then recheck the row to trigger regeneration. Or alternately, and perhaps more ideally, allow the user to add/remove/manage sentences from an explicit/visible "Generation Queue" akin to a tool like Handbrake.
• Store window and font sizes in a state file when they are changed so that they can return to those defaults when starting.
• Add (optional?) column to Sentence table that displays the speaker name instead of relying on color association. Row coloration makes it hard to determine row selection, and places a memory burden on the operator to remember the color associations.
• Either co-locate TTS Engine and S2S Engine dropdowns to Right Column above the engine's options, or move them to an overall configuration menu.
• Book Name in left column can be combined with the Top Bar that displays the book name.