Are query results 50% better with head titles USED WITH meta keywords?

Does anyone think query relevance would be increased, say 50% with
S.E. robots looking for meta HEAD TITLE as well as mere meta KEYWORDS
alone? To me seems like everything revolves around a subject, but with
KWs alone, has gota be a bit irrelevant. Also, the deep invisible
web has no self-index where I think subject and predicate (or head
title as 'subject' and keywords as 'predicates' (where the relevance of
mostly repeated pattern of connection occurs both S + P) fits this sort
of criteria. An auto-index is needed for the ever-changing world,
because it seems there is nothing wrong with any current S.E.results,
just that the user does not know what he's looking for.

And perhaps maybe sometime in the future he or she will be able to
'compare' the query to result in an overview (category) like fashion
where the user would type the query 'thesis' (sub + predicates) (same
as head title + kw's)), and the computer 'report' would: 1) 'compare':
Get other heads by linking the thesis pred(s) as kw's). 2)
'Categorize (sourceout) the source(s)': by making both the 'thesis
subject' and all the other 'co-subjects' the predicates (kw's) of the
source (head). And finally, 3) 'Verify' ('match'): the predicate
(kw's) from the source (head) to the thesis's predicate (kw's), thus
proving the query 'thesis' "True" or "False".