A real "quick and dirty" feasibility look - probably riddled with errors, and possibly based on an incorrect interpretation of the invention:

To lower the pitch by one semitone, you must reduce the frequency to about 94% of the original frequency. Since frequency is proportional to the square root of the tension, the tension must be reduced to about 89% of its original value.

A typical piano string is at about 160 pounds tension, so the change in tension for a semitone change in pitch is about 17.6 pounds. Most useful core wires will have a yield strength of at least 120,000 psi, and will be "tuned" at 80 or 90% of yield - or at around 96,000 to 110,000 psi, so for a rough calculation, a core wire area of something like 0.00145 square inches is likely. (This give a core wire diameter of about 0.04 inches - possibly good for at least some of the strings.)

An 18 pound change in tension changes the stress level for the hypothetical string by 18/0.00145 = 12,414 psi. Since the "elasticity" of a steel (or bronze) core wire is on the order of 30 million psi for each inch/inch of stretch, the change in length is on the order of 12,414/30,000,000 or 0.000413800 inches "stretch" per inch of original string length.

The thermal expansion coefficient of steel is on the order of 0.000012 inches/inch for each degree F. Dividing 0.0004138/0.000012 gives the temperature change needed, or 34.5 degrees F.

You would probably want the "in-tune" temperature higher than the warmest ambient in which you want the tuner to work, so you would likely choose a "design temperature" of at least 120 F - in a 95 F ambient. It should be noted that both the strings and the soundboard will expand and contract more or less in the same direction, so the amount of tuning required would not vary too much if the room temperature changes.

It would also be possible to heat only the "short end" between the end anchor and the bridge to change tension on the whole string. This would avoid burning hammers, but would require a larger temperature change (because of the shorter length affected).

The biggest drawback is that warming each string by 15 to 30 degrees F would likely require at least a couple of watts per string. Because the resistance of the wires is low, you would need a low voltage but high current power source. With 230 strings in a typical piano, you might have to have a portable arc welder for a tuner power supply. Of course the resistance of the wires goes up quite rapidly once they start to glow...

Altogether - it seems like something that could be built. Why is another question.

John