[xp84]

I think a probe would be really great, because knowing the surface temp doesn't tell you how hot the thickest part of the food is. Although GP has an interesting point that if you cycle it for long enough, I guess the heat should approach equilibrium and the surface and insides should match.

[crazygringo]

Yup that's exactly the idea, for it to approach equilibrium.

And the microwave should be able to deduce how close it is by the rate of cooling of non-hotspots while the cycle is off. If it's cooling quickly, the inside is still frozen. If it isn't cooling at all, the inside is nice and warm.