I just thought I'd report success in a making a super high locking torque 'Hybrid' freewheel with dual row caged bearings. This FW has six synchronously engaging pawls (the largest factory offering being three synchronously engaging pawls)
Recipe:
1 x Tensile '96 click' FW inner
1 x ACS claws FW outer
(Yes you will have to buy both freewheels whole to cannibalise the needed parts from both, But the ACS one won't cost you much)
+= Hybrid Vigor
The two odd halves from different manufacturers mate together perfectly.
So why does this 'Hybrid' have such high locking strength? In standard trim, the 6 pawls of the Tensile FW inner mesh with the associated 32 tooth outer in 3x asnynchronously timed pairs (two at a time 'lock-in', 3*32 = 96 'clicks'). By substituting the FW outer for one with a tooth count that is evenly divisible by three, all 6 pawls engage synchronously, rather than in offset pairs. The ACS Claws outer's have 18 teeth and the pawls fit in the tooth valleys more 'snuggly' than they do even in the tensile FW outer. I suspect this mod would work just as well with an ACS Crossfire outer (30 teeth), but I haven't tried it.
Note: Whilst the ACS Claws FW's are snynonomous with poor quality, it's the inner FW component (that is discarded in this mod) that has the poorest tolerances. The outer section is more acceptably finished, and with the CNC accuracy of the Tensile inner, all 6 pawls engage at EXACTLY the same time. Try as I might, I cannot simulate a 'race' condition (less than all 6 pawls engaging in synchronisation) by turning the FW outer with even, controlled slow rotation to test for this fault.