bill - as always your most accurate source of information is Gary at CSS.
I don't know the driver models that you reference, but I do know that the CS3.6 midrange used Thiel's patented dual cone diaphragm which has its unique dynamic structure. Its dual cone stiffens the rim while reducing mass - a seeming contradiction, but true. Any other driver will have a different response profile and therefore none of the (quite complex) XO control circuitry will do what is needed. A substitute driver will work in a crude sense, but all elements of its own output plus its interactions with the woofer and tweeter will be incorrect.