They use the resistance wire as temperature sensor. Not being possible to use Seebek effect due to compatibility issues, or the actual bridge configuration, I indeed believe they interrupt the power to the resistance wire in the atomiser periodically, and they measure its resistance to determine the temperature closely. Initially only a wire with high resisitivity variation has been used (Ni200) but if they use a high-gain, high CMRR front-end (a good differential amplifier) any wire can be used for sensing. I believe the problem they face is not amplifier gain or noise immunity, but linearity and repeatibility. In order to get a reasonable linear feedback, they need to isolate, and quickly, the power source, controlled by a MOSFET usually. The technology actually exists under the form of "bridge motor controller" but the latter are relatively big in size and doesn't foresee an high impedance status, which is needed here for a good temperature control. As usual, it is matter of investment in a technical research that will be cloned the day after it hits the markets...
Way outta left field...but damn! I'll take your word for it...