Well as I said, for myself the applications I develop are limited by what I feel are the core principles of the kata and I get to these by applying a few rules to my analysis which weed out a lot of the other applications I'm presented with. I end up with relatively few applications that I'm happy with this way but the context of the technique is always kept in mind. Sometimes overlap does occur, but it is justified because all the techniques involved adhere to the rules.
So for example a sequence that goes: Punch, turn and block
I may interpret as punch (KO opponent) go onto something else, AND strike and grab then throw.
This works because App 1 may not KO the opponent so you move into App 2.
Turns to me are natural breaks in kata sequences. They can indicate throws but interpreting every turn as a throw I think is an error. Sometimes it is what it is.
In Heian Shodan (Pinan nidan) the first 270degree turn into low block is clearly a throw. The second 270 turn, less so. Heians are especially good for this kind of contextual study because as well as moves being taken in sequence there is also 4 other kata to consider. The type of throw that could be indicated in the second 270 turn of H1 is much more explicitly shown in H3, so why force fit it into place when there is plenty other stuff going on in the kata. Not to mention that the Heian were never meant as more than basic self defense, IMO they are the Okinawan equivalent of "Street Smart Self Defense in 6 weeks".
Incidentally many Kata analyses seem to focus on throwing the opponent to the floor after he's been hit. I say just hit him again until he falls to the floor without assistance. Contrary to popular belief this can and does happen and is usually much easier than throwing someone (hence more often than not where I wrote throw above I'd have actually decided on a joint lock, not to restrain, but to open for more striking).
It's Shotokan not Shoto-can't!!!