IainGrace, 41d does breach the rules, but I suppose that as the indirectly indicated elements and the definition are so transparent the editors decided to allow it. I hope it's not the thin end of a potentially troublesome wedge.
I'm not convinced that 34a is kosher either. In the past the editors would have required an additional operator.
I see that solvers are a bit split on this, but that's normally the case with Waterloo. I didn't think it one of Waterloo's best but it was quite good fun and I was just grateful that it didn't take more than part of yesterday afternoon to complete it, following last week's marathon. I agree that some of the titles are very contrived and could have been better. It took me a while to decide on the entry for 9a. There were about three likely possibilities, but initially I could think of only two titles using two of them, neither of which could be reworked using the number of letter given. It didn't help that I found a Chambers word for tailor that could have been a corruption of two words, and that sent me up a blind alley. It strikes me as a bit unfair. If a solver doesn't know the work, or cannot find it despite extensive searching, I don't see how it's possible to determine the entry.