You’re right. I may have made a wrong deduction during my research. That was an inference I drew, not a confirmed fact. Sorry for the confusion!
What does seem supported by primary sources is that Mitsui’s CNT pellicle program targets ≥92% EUV transmittance (revised down from their original ≥94% target) while Canatu’s FC-CVD process achieves >97%. Whether Mitsui manufactures their own membranes or sources them externally, that transmittance gap appears structural and significant.
So a partnership with Canatu could be a good deal for them in my (not expert) opinion.