I Thought I'd finished this puzzle on Saturday.
On checking my grid for any errors I noticed that there's an ambiguity in the entry of 33 down, according to Chambers, which has been spotted above. I decided that there must be a way of resolving this ambiguity thematically and therefore concluded that an earlier assumption was wrong. When I checked all the letter changes in the p(url) cells I expected to find the letters had moved elsewhere, but this is not the case. For instance, the first letter of the answer to 10 down does not appear anywhere in the final grid.
IF the rational for the final grid is that, after determining most cells according to a pattern, the remaining 20 or so cells (ignoring the thematic entries) are determined by the only things that fit, then there's an ambiguity in 33 down. BOTH possibilities are certainly listed in the Chambers CD Rom, and we've already had at least 2 cases this year where a plural form has been used that has been implied rather than explicitly stated in Chambers. If you have the Chambers CD ROM, enter the first 4 letters that are definite followed by '?' for the unknown final letter. You'll get 2 answers.