It can be. There’s other methods too. Newton’s Method (nearest squares) is a pixelating approach approximating rough solutions smaller and smaller ’til you get “close enough”. Yours is kind of a chipping away marble way which also works.
There’s also a whole methodology of invention that a Russian dude came up with called TRIZ.
That can be a neverending task in itself.
Law of diminishing returns.
I can undercut almost any argument and I bet you can too. I think for good ideas you have to hit a point where you plunge ahead and just do it. Get rid of the big glaring obvious flaws and then run with it.