Hi Nicolas,
On the old EPSA cases, instead of a screw-in caseback, the caseback was an approx. 1/4 turn to lock down on the o-ring. Three slots on the caseback engage with 3 tabs on the case itself.
Concering the pinion: One of the drawbacks of the old cases was that the depth of the crystal as loaded into the case determines how much float the internal bezel has. The problem as I understand it, with this arrangement is that too much depth can put radial load on the shaft of the upper (2 o'clock) crown. I think this is why so many compressors have an incorrect 2 o'clock crown - the shaft has broken and the crown has been lost (pure speculation). It would be interested to see what IWC have done here.
These old cases are fascinating (after all, look who used them...) but information is scarce. Crystals are as difficult to find as orginal hash-marked crowns, so I dare not delve too far into my watches to understand futher.
All the above wouid be instantly answered if a cross-section of the watch is available. Have you seen any?
(by the way, I agree with the Green Lume on the IWC, it's just that artificial patina also detracts from a watch, particularly a watch like a Polaris that is near, in my eyes, to perfect: The right amount of complication (actually useful: alarm and elapsed time), the water-resistance, the legibility, the history...etc)
Thanks again,
BDLJ