Ah, now the real requirement surfaces. Largeness isn't your concern it is efficiency, right?
I see three * analogRead in your loop plus some delays. I would be worried about those myself. Each analogRead takes 104 uS so you are up to 312 uS straight away. And then you throw in a 100 uS delay. If you want responsiveness you need to rewrite without delay, and do the analogRead asynchronously.