Many thanks for all the helpful comments: as to method of measurement, I put a 3MT precision lathe test bar (brand new from LPR Toolmakers) in the SX3's 3MT spindle, and rotated it (in tapping mode), against a DTI; which showed 3 thou runout. I know the 3MT test bar is OK, because I recently had my Hercus Model C lathe (a clone of the South Bend Model C) restored by F. W. Hercus, here in Adelaide, and no run out is detectable in the Hercus (3MT) headstock.
When the weather is cooler, I will repeat the test using an Allen key to rotate the spindle by means of the drawbar, instead. Also check whether the spindle has any detectable side play in the bearings.
As regards the table, I mounted the DTI in the SX3's drill chuck and wound the table from full left to full right; it is flat to better than a thou from LHS to about 3/4 of the way over to the RHS, where it commences to rise to 5 thou at the extreme RHS; so I don't think it's a tramming issue. Again, when it's cooler, I will repeat the test using a 3MT/ER40 chuck/collet to hold the DTI, instead.
I do have a copy of Dr Georg Schlesinger's "Testing Machine Tools", sixth edition.
I should also add that my SX3 is exactly 'as delivered'; ie I have not performed the type of work recommended by arceurotrade; who incidentally have been most helpful; although I purchased my SX3 from HAFCO, in Sydney, via General Tools in Adelaide (because I had no facilities to unload a 200kg crate). General Tools very kindly used an engine hoist to put it on my bench.
I did my machine apprenticeship at English Electrics Willans Works, Rugby, in the mid fifties; I well remember that the castings for the Fullagar Diesel engines (intended for ships, submarines, locomotives and power stations) were left outside to weather for six months, before any machining. So, I believe that my SX3 table may well have warped after machining (it has no evidence of 'hand scraping' marks, only surface grinding).
Euan