I'm grateful for all your thoughts there. They've definitely helped my thinking.
Like Big Monk says, I'd like to try for a bit more control on this one. Without a lab and a white coat, I know it's not going to be precise, but I have had a couple or fairly flakey carbonation phases with tripels before, despite carefully measured priming sugar, so I'm willing to give something a go... I think.
Correct me if any of these assumptions are wrong, but I've read that 1m cells/ml is roughly the limit that we notice a haze in (otherwise clear) beer, viewed through a test tube. That is also the lower end of what Stan H tells us the Belgians are happy with for carbonation, in BLAM. So I'm thinking I will lager the beer down a little longer than usual, enough to get it around that threshold, as best I can tell, and follow a) with around 2g of dry yeast.
Denny mentioned s-04; luckily, I have a sachet of that in the fridge, and it certainly sticks like s*%t to the bottom of bottles, and will hopefully be a good choice. I was unlikely to use it for anything else soon, either! Assuming 6B cells per g (from Fermentis), which is a conservative estimate going by some reports, then that 2g s-04 gives me just over half of my million cells per ml in the 21L (max) I am likely to get into the bottling bucket. It's a finger-in-the-air really, but I'm going to guess that a week or two at 37f is still going to leave enough of the primary stuff in there to make up any shortfall, and I don't want excess sediment, as per the title.
One other thought, I guess I'm going to have to let my chilled batch come back up to room temperature for priming and bottling, to avoid shocking the freshly rehydrated s-04. I'd usually just bottle it cold. Or am I better to rehydrate, and then gently chill down the s-04 mixture... Pretty hard to be gentle and accurate with just 20-30ml of liquid!