Personally I'd want to be quantitative about the stimulus that is hypothesised to be necessary or relevant here. If current in coils is relevant then all we have is magnetic fields, and maybe some induced electric field due to current dI/dt if Triac controlled. These induced electric fields don't look too high for me because the primary is multi-turn and has voltages (and therefore dI/dt) limited by Triac breakdown. Further, the magnetic field is pretty low. For example; 30 turn, 5cm length, 1cm, 100A, gives only 0.07T.
What fast edges will do is give you EMC issues in any thermocouple sense amplifiers which can be difficult to track down.
In my reactor design the control and test port thermocouples both sit right inside the central flux of the heater solenoid coils. The TC's are connected to amplifiers/interface boards of several kinds during different tests and during calibration. At various times the coils have been supplied with triac-controlled grid power, 15kHz DC, or square-wave 'pseudo-AC' at many different frequencies. The heater coils are similar to the ones you describe, but because they are larger 21mm diameter, and the current max is around 15A RMS the field strength will only be a fraction of your 0.07T.
However, Much to my initial surprise I have never seen any induction effects or similar monkey business affecting the TC output. The only thing they hate is when the outer woven mesh screens on the leads touch. Then they become very erratic.
