I watched the carb temperature all winter and never once saw the temperature get down into the icing zone, even during an idle power descent on a cold day. Enough heat is transferred from the oil sump to keep the carb warm. So I'm confident that my particular engine just doesn't need carb heat on descent. (Of course it's there if I ever need it.) I'm still debating the wisdom of removing the temp sensor that sticks out into the throat of the carb. Several commenters have recommended that and they're probably right. What I have found over the last six months is that the rearmost cylinders 3 & 4 run lean at WOT on takeoff with the mixture full rich. That's probably not a good thing, but to compensate, I pull back on the throttle a little when reaching 500' and the EGTs even out. You can see that in the graph below. It's curious because I would have expected that with the throttle fly vertical, more fuel would reach the rear cylinders than when the fly was aimed more toward the front. I think my next step will be to drill out the "peppermill" holes I drilled in the nozzle years ago to provide more fuel to the rear cylinders. IIRC, they're only #57 now, and in the photos of the Marvel nozzles, they look much bigger. "It's Experimental"
Full Throttle Throttle back Leaned out for Cruise