I have to admit after working with the concept for awhile, that for proofing, numbered words does has some use, especially in clarity. For instance, may I propose a shorthand? Something like
2r.8.8 <ee>=<e,e> or 2r.8.8 {C}={c,c}.
This would be short for "At Folio 2r, line 8, word 8, you have EVA <ee> written as one glyph (also written as VGBT {C} ), when I think this should be EVA <e,e> as two glyphs, (which can also be written as VGBT {c,c}.)
Either proofing language works as long as they're identified with <> for EVA and {} for VGBT, although I expect we'll probably be using EVA for most of the discussion. This might be a good way of submitting long lists of proofing corrections and an easy way to maintain them in a list for comparison?