Given that the guy I "ken splained" to didn't know the difference between an R-22D and an R-22S and why one isn't suitable for diving applications, yeah I figured being a textile engineer didn't matter.
Right, the "R22S" O2 sensors are the same as "R22D" with the exception of the removal of the hydrophobic membrane over the sensor cathode, as well as the removal of the humiseal treatment of the sensor's printed circuit board.
That implies that the "S" sensors won't work reliably in a CCR application. However in practice a number of divers, in a number of different units, seem to be having great results. Meanwhile, AI's cells are better than they used to be, but they still have a fair percentage of cells that are DOA or are marginal performers.
You're basically balancing one problem against the other, and looking for the best balance of performance in a specific application. That's more or less what engineers do - they apply technology and they don't necessarily design something to be stronger, or more effective, but rather they design it to do the required job while being lighter, easier to produce, or less expensive.
I recall a discussion of when the first Russian defected in a MiG 25. Much was made in the press of their use of steel in the airframe, implying it was overly heavy and substandard to western technology. Meanwhile, in the back room, the western engineers were marveling at the Soviet skill in producing and utilizing thin stainless steel structures that were lighter and stronger than aluminum structures in high heat applications (not to mention milled titanium fittings that were beyond the ability of western aircraft manufacturing at that time). I got that information from a professor/former Navy aeronautical engineer in my first year of aeronautical engineering. The message he was preaching was what works is what works.