Almico wrote: If it was measuring bean temp so accurately, why the need for algorithms to interpret the data?
I agree with most of your sentiment. I think the problem is the fact that they have to have an algorithm to remove the error induced by the vanes of the drum that will have an entirely different emissivity than the beans. By the time the Alillio is broken in, the drum shouldn't change too much in terms of baked on shiny black oils, so hopefully that won't be an ongoing problem.
I would like to know if they took some chances and placed their newer sensor in other locations. It seems if there's a small fan keeping debris away from the sensor, you might be able to get away with mounting it in other locations, where the vanes aren't in the way (which would have to be down towards the bottom front of the drum by the door.
Something could likely be rigged onto a makeshift replacement bean tryer that aims right at the spinning bean mass, but in front of any vanes.