Personally I don’t have a problem with 25ac: I think compilers feel entitled in “hidden word” clues to use as much surrounding verbiage (within reason) as is necessary for the sake of the surface (and Gordius was presumably going for a Disneyesque surface that wouldn’t have worked without “in film”). If FAMOUS is in “oF A MOUSe” then it’s also in “oF A MOUSe in film”; it’s merely a question of how much verbiage goes unused. I’ve seen much longer unclued fodder in this type of clue.

I also wrote in COAL – the clue, to me, points more that way, but is perfectly fair IMO. DEAD HEAT easily forced the rewrite. I also see no need to capitalize “just in” – surely setters are allowed to abuse capitalization, just like punctuation, to build a tricky clue? I also liked that clue a lot.

I agree “in film” is redundant. I wonder if it was part of a different way to build the clue that went undeleted by accident? Like “Part of a mouse in film”?

Thanks for the blog, UY, and a fun exercise, Gordius!

Some obscure words indeed (WERRITTY, MISPRISION) but they were perfectly gettable from the construction.
In 4d we were very careless, entering DISCRETE instead of DISCREET [not a very good clue anyway, because 'its creed' echoes quite clearly (though in the distance) 'discreet']. As a result we couldn’t find ATRIAL (17ac) which was however in hindsight a nice clue.
We weren’t taken the double crap in 3d, but thought just like others that KNOT (24ac) was excellent.
That said, just like quite a few words in the SE, it fell rather late in place.
Reason for that, of course I would say, the COAL/KOHL dilemma in 24d.

Gervase @8 says: “I spotted the homophone at 24d immediately but read the clue the wrong way round and inserted COAL at first, which rather banjaxed the SE corner until I realised my mistake”.
Your mistake?
We think the clue is clearly pointing in the direction of KOHL (although COAL might be justifiable).
One that tupu @14 liked was 21d (despite indeed being a chestnut), but we didn’t like this one at all because this is another example of an ambiguous clue. I know it’s a “cross”word, but when looking at it as a stand-alone clue, the solution could just as easily have been LAMINA.

A nice twist in TIMBERLAKE (23ac), original too. But it not fair to nót capitalise “just” (and also against the rules). It must have been possible to do something about that without ruining this great idea.

Finally, 8d (ENTREE) is a really nice clue, but I have my doubts about the use of that little word “of” here. I know, it’s just a linking word, often used by other setters including Araucaria.
Gordius uses it here as: “construction” OF “definition”, while IMO this is less acceptable than: “definition” OF “construction” (which is perfectly all right).
Eventually I cannot be bothered too much because I know setters do it like that, but my occasional alter ego Dalibor made clear that he ‘would keep far from it’.

But overall, a well-spent little hour.

I found this much harder than usual for Gordius, but very satisfying to solve.

Thank you for the blog UY