I think it also shows that breakthroughs are not driven by innovative and research but mostly by scaling.

If this is the case, makes sense that frontier labs with similar access to compute driven by funding on same order of scale can produce improvement largely on similar pace