I suspect the answers are going to be "how long is a piece of string?", however - here goes. Once the idea that HST sets were "married for life" was abandoned, how fluid were the power car pairings? Other than failures, were power cars swapped out for any other reason than Exams - and at what sort of intervals?
Put it another way: If power car A was observed paired with power car B at 09.00 and again at 18.00 the same day, it's as near as possible a cast-iron certainty that they were also paired at 14.00 that day. Guessing also that if they were observed paired on May 19th and 21st, it's pretty certain that they were together on the 20th - how far could you push that assumption?
Put it another way: If power car A was observed paired with power car B at 09.00 and again at 18.00 the same day, it's as near as possible a cast-iron certainty that they were also paired at 14.00 that day. Guessing also that if they were observed paired on May 19th and 21st, it's pretty certain that they were together on the 20th - how far could you push that assumption?