"Counting frames will not give you the timing being used."
Yes it most certainly will!
It will give you the relative timing of the actions with far higher precision than you could ever hope to hear. And that enough is to prove that your arguments about when the steps fall are wrong, since your theory has different amounts of time between the steps than actual measurments of leading dancers reveal.
To determine where in the sequence the beats actually fall, take a noncontroversial relationship, such as that step 2 lands on beat three. From that you can then work forwards and backwards to determine where in the music everything else falls. And this is actually a lot more accurate than listening to the audio, because you can't be sure that the audio and video haven't shifted relative to each other during processing.
"You will need to explain a bit better than that. How do you propose to count 4.5"
"4.5" is the same thing as "the and after four" Maybe where you are a decimal fraction is indicated as 4,5 ?
"To come more into the real world. If you had the same tape i have you would see the right foot arriving one hair breadth before beat two."
That's exactly what I documented. Your error is in thinking that the third step lands on beat four, that is WAY, WAY, WAY TOO EARLY. The third step's early limit is beat 4.5 - which is to say, the and after four. And on a dancer who can draw things out like Sinkinson it will be more like 3/4 of a beat after beat four.
"What you didn't take into account was what was happening on beat one with the LF. You silly boy"
THE LEFT FOOT DOES NOT STEP ON BEAT ONE! That's your observational mistake, instead it steps somewhere between beat 4.5 and 4.75 - a half to a quarter of a beat before beat one. Your real error though is in somehow coming up with a feather step that lasts less than a measure. WRONG, WRONG, WRONG. Each SQQ figure gets an entire FOUR BEATS WORTH OF TIME - but the catch is, that period of time does not line up with the four beats of one measure. Instead, it starts a bit after the start of the measure, and carries over precisely the same amount into the next measure, so that it's exactly one measure worth of time, only offset from the barline. Each SQQ figure has the same basic timing, each is 4 beats worth of time, so in a series of them the offset from the measure is constant.
Go actually measure Sinkinson on the 98 blackpool tape you mentioned having, and your mistakes will be revealed.