What top "numerology" theories can't seem to do and where are the alternatives

Carl, here's a question for you and several others in the lieu of Igs's doubts about so much "numerology" floating around that doesn't seem to match what his ears hear a fair deal of the time...

   Do you think JI dyads always work on a basis of the higher the Tenney Height, the worse the concordance?

   As counter examples...do you think 16/11 sounds better than 22/15 or 13/11 better than 18/13 or 11/8 better than 13/7?

  Of course I'm thinking you don't...you said yourself Tenney Height loses accuracy over 70 or so...but then again considering Harmonic Entropy is based on Tenney Height.  But concerning odd limit...13/7 also is rated as worse than 11/8 and 22/15 worse than 16/11.

  It's not that these theories don't to anything...it's that there are some gaping holes in all of them over 7-limit or so...and nothing in the "booklet of consonance theories" seems to handle any of these in detail or reliably.
And well many of the methods here are fine for many people, often whose idea of a good chord is one without a single 11-limit ratio, it really comes across as a bizarre constraint to people like myself and Igs. People who find, by ear, lots of higher limit chords quite acceptable and need a good alternative to navigate through them with.
-----------------------------
   Second point...why does everyone seem so stuck on using perfect 5ths?  Sure, they are the best when you can get them...but what research is being done about using "semi-pure" 5ths like 14/9?   We seem to have a self-fulfilling prophecy here and pointing to perfect 5th based scales and, the more we "succeed" at this...the closer we get back to mean-tone and ultimately 12TET.
------------------

  Also, I'm interested...what do you think of clustered triads (IE triads with end bounds of no greater than 5/4 or so)?   

   It's a huge thrust of what I'm interested in and yet no one seems to be able (or just wanting?) to consider them.  Even using Adaptive JI to retune clusters in 12TET...I trust it could make many "used to be wrong" chords sound good enough for musicians to try them and eventually get more interested in microtonal scales with lots of such small intervals...

  How about 18:19:20 (yes, there's a 10/9 in there) or 18:20:21 (with a 7/6 and a 10/9) or 18:21:22 (with a 7/6 and 11/9)?  Or more tame ones like 15:16:18 (with a 9/8 and a 6/5) or 9:10:11 or 10:11:12? 
  Considering people on this list are so interested in some of the atonal music floating around...I can't see why it could hurt to, say, work on some tricks to make clustered chords a part of common music theory and not just a way to make "neighboring tones"....