well ...
On those few occasions where I've created a spread, it was part intuition and part study of the shape and form.
There's no reason why intuition could not suffise, as it seems widely accepted to draw and read single cards per question - whether the question asked is a "vague starter" or a request for further elaboration on the cards already drawn and laid out.
Now, that I think of it - most spreads seem to function as a helping hand if one doesn't feel to sure what the next card drawn should stand for ... or to add to the way a card is read: remember that a card's reading/interpretation is based on the card, it's position in the spread, the question asked, ... so "fixed" spreads offer one extra "grip" on the card based on its position in the spread.
I've come to the point where I frequently use four spreads, two traditional "classics" and two I've created myself ... but my use of those sometimes has more to do with my need for that extra "tip" on how to read the card.
I do know that some readers (and of course certain people on this forum will feel targetted, whereas I don't mean anyone in particular) use spreads because it makes the right impression on querents - it comes across as "the professional thing to do" and makes the whole experience more believable for the doubting querent.
If you're looking for a spread that suits you in your way of reading, why not start with a small existing spread - and add on to it in a way that answers your specific need for further clarification or insight.
All things said and done, a lot of spread with many cards use only small variations on the meaning of each position (and yes - I'm guilty of that myself in my recent addition the the Ankh-spread): positive and negative influences, past-present-future influences or forces, warnings and options, possible outcomes, ... the shape of the spread doesn't always seem to make that big a difference.
PS/ and if that doesn't get me a whole barrage/discussion of disagreements ... I don't know what will - devlish grin.