This is a very old idea I had in the seventies when at college, and I have just tested it.
When you put a toaster down, the element radiates heat. The current rises until the heat radiating from the toaster equals the heat being absorbed by the toast, the heat shining out of the top of the toaster,
and all the various other heat losses.
The toast and the other parts of the toaster reflects some of the radiation. They will radiate a tiny fraction but this radiant power goes roughly as the 4th power of the temperature, so we can probably ignore the re-radiation of the toast itself.
Most white objects are white in the infra-red. Most black objects are black in the infra-red. Unless bread or toast has an unusual reflection spectrum, this would mean as it started to go brown, the toast would absorb more infra-red, and the toaster element might get a little cooler, which would change its resistance so it takes slightly more current. This rise in current ought to happen when the bred starts absorbing radiation, so it should work for all sorts of bread (I hadn't got any brown, but I will try that on the next loaf).
I tried this. I had something that measured the current going to the toaster. I popped in two slices and I wrote down the current at 10 second intervals. The current went up to 3.73A, then dropped to 3.70A. It slowly went down to 3.692A and then rose to 3.712A. At that point it popped up a nice shade of brown.
I think the rise in current is too small to say it works. I could have kept going until it burnt. However, I think the reason the rise in temperature is so hard to detect is from the toaster design. Energy is cheap for a toaster that runs for 150 seconds at about 800W. You could make a more efficient toaster with a reflective inside and lid, but it is better to allow convection and use smoke as a warning. Probably very little of the 800W is actually going into the toast, and the small change is hard to see. The power supply is voltage limited so if the toast absorbs more radiation, the element may become cooler, but its resistance drops and it passes more current and the power output rises. If we had a current limited power supply we might get more signal.