partial progress: I still can't get it to idle but I have learned a bit more about how the carburettor works
After realising my mistake with the petcock I thought I had cracked it, but it was showing symptoms that I think indicate the idle mixture was too rich (I could start it without choke and the plugs were covered in dry soot and - after a few minutes running - this sooty covering built up and made it difficult to start).
I took the carburettor off and, on closer inspection, I can see that the mixture screw is in fact an air screw that restricts air entering from an air way on the air filter side.
.. what's more when I squirted carb cleaner through the air way it was blocked, "bingo!, that would explain it why it was rich when at idle", or so I thought. Having cleared the passage way the engine now starts first kick and seems to idle better at 1 1/4 turns of the air screw but - unless I use the throttle - it stalls no matter what adjustments I make to the idle screw or the throttle stop and the (new) spark plugs came out covered in soot again. Although I have not ridden it, it does rev freely which makes me think it is a problem with the idle circuit although I must admit that I have reached this conclusion based on a fairly sketch understanding of how it all works!
I have swapped back the old air screw, which has a very slightly different profile to the replacement (the air screw and throttle stop screw were the only brass parts I used from my keyster kit), but this does not appear to have made a difference. I noticed from various parts diagrams that the airscrew was not intended to be used with an o-ring, and indeed the keyster o-ring I had used clearly did not fit properly and had been mangled when I screwed it fully home, so I removed it (I suppose this might have stopped the screw seating properly messing up the 1 1/4 turn setting, although presumably this would have tended to make it more lean rather than rich). Anyhow, this made no difference either.
I checked the idle jet was clean and the correct size (35 - genuine keihin part).
The other thought I had was perhaps I simply have the wrong carb. It seems likely it is from a different bike as it had the wrong throttle cable installed, but the body is stamped CA95 and I have seen pictures of the same model on other bikes. There were two models used, the PW20HOV which is the earlier model and did not use power jets, and the model I have which did (I think Honda called this the PW20HA3). The service manual does not give separate setting instructions for the two versions, so I suppose it is possible the settings for the newer carb have been omitted, however the only other CA95 instructions I can find (on a service bulletin from 1973) say the float should be 19.5mm from the underside of the body and the air screw turned out 1 3/8th. It won't idle at all with the screw turned out this far and stalls if i open the throttle, and the float height when measured from the bowl makes no discernible difference.
the air filter is NOS.
The choke opens and closes fully and smoothly.
So I am a bit stuck. The only thing I can think is the fuel level is too high in the carb. I did notice the upper part of the carb body is warped (you might be able to see the slight curve to the edge in the picture above). This does not affect the seal with the fuel bowl, but I suppose might mean my float height measurement is out by a mm or two. Could this be an explanation or am I clutching at straws? Could there be another explanation for these symptoms? Questions questions questions - sorry!
PS Although not relevant to my investigations this might be of interest to other CA95 owners: I found out there is a misprint in the manual for this carb: it says the power jet and power air jet are both 130s, in fact the power air jet is 130 but the power jet (the one behind the cap screw) is 180 - this is the configuration that was in my carb and has been confirmed by others. There is some debate about whether the power air jet should be open to fresh air or connect to the air cleaner, but I am 99% sure Honda meant for it to get filtered air because they made a
air cleaner with two inlets and it does not seem possible to explain the provision of the second connection otherwise. I think the confusion arises because the single hole filters - intended for the older carbs - are still available.