barretter I am in a similar position. I've solved all the clues (though unsure on parsing the Rossini one) and have a 90% filled grid, most of which seems sound ... but I have 2 across and 6 down clues unused, and I can't squish them in without causing more clashes than I assume there should be. Something is wrong somewhere because I don't have full symmetry. My bottom four rows are intact and I can see the second part of the quotation ...
If you look up the first 5 letters of the Rossini clue I think all will become clear. It may be that I have the grid co-ordinates wrong but it seems that neither of the ND across solutions will fit in the available space but if you change its first letter you get a couple of words that might satisfy the two defintions. On the other hand, apart from its first letter I can't make EI down fit at all so far. Its last four letters form part of the name of one of the "characters", I thnk.
Hmm, I have the ND (6) clue in with no clashes, but three other across clues (including ND (12)) and six down clues not in at all. Think I might need to start another draft (after a lie-down).
I can only find one 7 letter Scottish word which fufills the criteria, presuming the name to be LS, but it appears to destroy the 12 letter word for moths which I have in the third row from the bottom, so do I have the y-axis wrong?
No that's where I have it. You seemed to be implying that the last four letters of the solution to NE had been replaced by 4 letters from the name of the author. If you just mean the 4th letter that seems to be OK but I still don't get the rationale for the suppression of the last three letters of NE and of the letters of other solutions.
The last 4 letters of NE and the first 2 letters of ND become the N of the surname. The replaced 6 letter jumble is one of the thematic names (2 words).