[38321003thrw]

I have a video link to an MIT lecture that disproves your assertion. Cache aware, cache adaptive, and cache oblivious algorithms are definitively analyzed using big O. In fact it is big O notions that provide assurances about these approaches.

Mechanical sympathy has little (if anything) to do with big-O, btw. LMAX, the poster child of mechanical sympathy, is really about minimizing the use of memory coherence machinery and not taking the substantial latency hits.