That doesn't make sense. A bi-metal switch is just that, a metal bridge across two contacts which bends when heated and breaks the circuit. So the switch itself cannot be 63.5 Ohms, you are measuring something else. When the knob is turned, a screw thread enforces different mechanical pressure to the bi-metal strip so it takes less or more heat to bend it far enough to open the switch, the electrical resistance should be close to zero when closed and open circuit when open.
It is just possible, but unlikely, that you have a timed thermostat. one which does NOT measure the temperature of the vessel at all but instead has it's own heater built inside it. These are very inaccurate as they can't monitor any external temperature and they work on the principle of cyclicaly turning on and off by themselves. When the knob is set to a higher temperature, they just spend longer 'on' and less time in the cycle 'off' ao the average heating power is higher.
Your power calculation makes sense but if the thermostat alone measures 63.5 Ohms, it means it alone is dissipating 750W which is extremely unlikely (I'm assuming the knob isn't glowing red hot !) although the main heater possibly could be that resistance. In any case, you should measure the resistance when it is at operating temperature, it will be considerably lower at room temperature.
Brian.