"Edition is missing here" needs more consistent application

[ebeshero]

C07 features some challenging revisions to track, because this is where
a) the MS (SG-A) witness first appears,
b) where the Thomas texts contains lots of passages marked for deletion as well as as side comments about those deletions (like, "bad"), and
c) the 1831 text makes some very long total overhauls that are totally different from the earlier material.

In studying the 1818 view of C07, I'm finding the long additions from 1831 easily enough, but it's not immediately clear that there is no corresponding 1831 text for long stretches of 1818. We should try to make it clearer where a given edition is missing or has totally replaced a long passage of text.

[ebeshero]

Given that C07 is is one of the most complicated passages in our variorum, we may need to do some hand adjustments and fine-tuning in units like this that we won't have to do in most other variorum chunks. But that's just a conjecture--if we can find a systematic way to indicate that a witness is missing from a given <app>, we should be doing that automatically if we can.

[ebeshero]

Or possibly where the passage missing is very long, there may be a way to flag a lengthy stretch where one witness is totally missing in a more "macro" way than we're currently doing.