Empirical question: Can any kind of cluster analysis result in identifying robust selected subject-domains/content-areas for search specialization, still allowi

Empirical question: Can any kind of cluster analysis result in identifying robust selected subject-domains/content-areas for search specialization, still allowing for good coverage with a set of these? This could fight what you identify as an exponentially difficult problem. It could also allow for selective marketing and mind-share.