Update - So my own design locking tool (6mm laser cut steel) worked perfectly and despite the newly arrived chinese castle nut socket turning out to be not really the right size (too big) I was able to improvise and get sufficient purchase to undo the nut. It wasnt very tight at all. I removed the flange, inspected (it looked fine), cleaned and regreased. I re-fitted nut and tightened as best as I could with a 1/2" drive ratchet and now there is none of the previous slop/backlash between flange and shaft. I re-staked the nut on both sides so hopefully it should be secure and remain tight.
What I cant understand is how such slop/looseness could have developed given that the nut was staked well before removal, albeit only at one of the two points? I cant believe that this looseness could have existed for an extended period without symptoms; I'm starting to think that perhaps the knocking noise that lead me to think it was a failing inner CV unit was coming from the loose drive flange.
I'll fit the new inner CV joint whilst the hub assembly is off the car, reassemble (with new cap head screws and nylocs) and see if (hope that) the knocking has gone.
Update - I put everything back together yesterday afternoon and have just completed a 20 mile test run - no issues and the knocking has gone, yippee!!
Many thanks Eric for the tip-off regarding the drive flange lock nut. I really dont want to believe that this was the cause of the noise but I suspect that it was! I'm still struggling to comprehend how given that the nut was staked albeit only one one side how it could be so loose. My theory is that the staking wasnt quite deep enough to be totally secure and some how the nut has moved under load and tentatively loosely re-locked itself 180 degrees back. After retightening the original point of staking on the nut coincided exactly with the shaft cut-outs. I just restaked and did the opposite side.
Did inner CV anyway (complete with new bolts and nylocs) and regreased outer CV.
Fingers crossed problem resolved .......
