You might want to save those old panels to play with someday
If you're handy and up for it, you could replace the spacer/adhesive tape and install new diaphragms. Whatever tape you need would be available from McMaster-Carr online. And the 12-micron Mylar C that ML uses is available on Ebay for about $35: Electrostatic Speaker Membrane Dupont Mylar C 12um 20m for sale online | eBay
Because ML panels are curved, the diaphragm must be tensioned predominantly in the height direction, as any significant lateral tension would pull the diaphragm into the rear stator. You would only apply the smallest amount of lateral tension needed to pull out any wrinkles in the film.
ML diaphragms are tensioned quite high, so they use the thicker 12-micron film to withstand the higher tension. The diaphragm's drum-head resonance (that sound you hear when you tap on the panel) is set by its tension.
I tension my diaphragms to achieve a drum-head resonance of about 100Hz, which equates to about 1% - 1.2% elongation, and I measure the elongation with a scale. But ML uses much higher tension and a thicker film so-- not the same. It's important to get the tension correct because ML will have tuned the crossover and notch or shelving filter circuitry to match a specific diaphragm tension/resonance.
Unfortunately I can't quantify for you the tension/measurement for a ML diaphragm. But if you have a good panel for reference, you could measure it's diaphragm's deflection using a weighted probe and scale-- and then tension your diaphragm to the same deflection.
The probe could be a small weight attached to a probe small enough to insert thru a stator hole, and soft enough on it's tip to not puncture the diaphragm. The weight could be stacked coins like quarters (whatever works). You would measure the deflection in the center area between the widest spacers.
I was told by an ESL builder on the DIY Audio Forum that ML tensions their diaphragms quite high (about 250Hz drum-head resonance) to enable using the "distributed resonance" technique to enhance the panel's mid-bass output.
ML's approach arranges the lateral spacers (spars) with sequentially decreasing spacing, which breaks up the single/loud drum-head resonance into multiple quieter peaks spread over a wider bandwidth.
This approach does enhance the mid-bass output but I'm not a fan of using the drum-head resonance because it isn't music.
I prefer to tension the diaphragm to 100Hz or lower resonance, and then avoid exciting it by crossing the panel out and the woofer in at least one octave above the resonance using a very steep crossover slope