Your experience with ice makers is totally different than mine. I
don't notice any difference in running time on mine when it's making
ice versus when it's not.

Noticing or not noticing isn't physics. One has to appreciate
thermodynamics and the heat of fusion to understand why ice making is so
much more energy intensive than making up heat losses through a well-
insulated cabinet or warm air infiltration.

Also, you vastly overestimate the sublimation effect.

I haven't actually made any specific estimate, over or otherwise, but I
suppose you mean to say that sublimation and the other wasteful phase
changes inside a refrigerator-freezer are trivial, when if fact they are
a major factor as evidenced by the need for frequent defrosting and the
pivotal role that plays in efficiency.

The trickiness of all debate over efficiency is that you can call things
like sublimation trivial that are indeed small, but then you're
comparing them to things that have been optimized down to very near
zero, like heat gain through insulation, which is what the DOE test
measures, then the efficiency ratings based on near-zero effects are
still completely spoiled by the comparatively large sublimation type
effects that aren't in the DOE tests.