My recollection is that string theory was welcomed as the solution to the problem in Feynmannology--that the two 'four point functions', one with a Feynmann diagram that looked like a Y joined to an upside-down Y, and the other which looked like that rotated 90 degrees had to count as the same rather than different--and so rapidly became dominant that it, DESPITE NOT YIELDING A TESTABLE PREDICTION FOR 40 YEARS, now is an impediment to any new ideas breaking out or being funded. I challenge you to find an American university where the physics department offers enough graduate courses to form the basis for a research program in theoretical particle physics on any ground other than string theory (or its even more nebulous daughter "M-theory"). The only real competition comes from European mathematics departments where Connes' recovery of the standard model from non-commutative geometry is actually studied. Your analysis is completely confused. String theory does NOT dominate particle theory, it has not been around for 40 years.
I'm afraid your statement about Feynman diagrams doesn't sound familiar.