Thanks all, OP here again! It does tickle me how threads develop a life of their own (and even have children, it would seem!).
Anyway, an update on progress. The new bearings went in the freezer (-18C apparently) for 24hrs, and the housing went on a radiator to warm up. The bearings pulled in easily, with only a minimum amount of persuasion required. Taking on board Neil Wyatt’s comment that his bearings maybe didn’t seat down fully initially (he used angle iron) I used a steel 60mm disc, with a 30mm (ish) nylon disc stuck on with double sided tape to keep everything aligned. Worked a treat.


The spindle also went in the freezer, and was easy to get through the bearings.
The metal spacer went on (mental note – a nasty fit, to be sorted). On my (later) mill this tube has small protrusions for speed monitoring (I’m not sure if these are magnets or merely reflective). Pulley, then the top ring nut.
With gentle and slow adjustment of the nut I could feel the free spin of the spindle reduce slightly – then I realised I had the nut upside down! Doh…
Then with the nut on the right way (and the mill re-assembled) I ran out of thread on top ring nut. This due I think to the ‘depressing nature’ (ha ha) of the angular contact bearings. There is simply more movement than in the original bearings.
I could make an additional simple washer type spacer but I was always going to make new pulleys for poly-v belts, so maybe I’ll plough on with that. And when I disassemble the spindle, I’ll face the ends of the steel spacer. I’m not convinced that anything is particularly accurate!
So, ongoing. BTW the bearings I’m using are SKF 7206 BE-2RZP – low friction seals. I only want to do the job once, and they should give years of good service.