Symetric "transconductanse" front end

[R Dijk]

Hi. Looking into symmetric designs as an approach for linnearity vvs gain. I was looking into different approaches David Hafler, Rotel, Slone, JC, Ti Kan and others...
It seems like it can bee divided into grups:
-gain vvs linearity.
-self biased/ biased input.
The dual current mirrors neds som modifications to work properly. But single mirrors in nonsymetric like Self are ok. Loking into R.Cordell he uses somthing like dual current sorce loading biased from VAS stage bias and then morroring.
National has some "front end chips" like LM49830 and ohters.
So there is a lot of good hard to beat symmetric amplifiers out there. Can we compeet with them?
So far i liked the JC best, i must admit, self biased linearity approach. But the mirroring was plundering in my head for a while. Then the idea of "half mirros" came up. VAS local feedback from vas combined with current sorce loading and the big ?.
Here is the front end , comment and modify !
Simulation results is a big smile so far.