For reference, I eventually isolated this problem: - I ran the engine
in the dark and observed plasma glow from two areas on the HT lead
connecting the coil to the distributor cap. I removed the lead and
stored it indoors for a week, re-connected it, and the car started
fine. The problem was resolved by replacing the lead, so my assumption
of electrics not being at fault was obviously incorrect. All I can
conclude is that the HT lead insulation was gradually absorbing
moisture from the air over several days and that the engine would
eventually start as the moisture slowly evaporated from the affected
areas due to arcing under the coil's HT during repeated cranking. The
car would start perfectly within two or three days because moisture in
air could not sufficiently penetrate the insulation over this period
to cause problems.