I think I will abstain from most of the major points, but just I will  
comment on number 4, in regards to the Walras that is referred to in  
Neo-Classical (and most other equilibrium analysis today).  
  
The story seems to me to be a very muddled one, with different strands going  
of in different directions, and I would be happy if anyone has corrections  
or additional points to add.  
  
In brief, starting with Marshall and Walras, they both had the notion of a  
'normal equilibrium', but where Walras used net excess demand (a la Quesnay  
I would suppose), Marshall looks at the problem through prices at demand and  
supply. Their conclusions are similar, equilibrium in exchange is achieved  
by price adjustments and equilibrium in production through quantity  
adjustments (noting that exchange happens through bartering traders with  
prices signals as the 'active force')  
  
In Pareto's Manuale, he also approached equilibrium analysis, but from a  
'tastes and obstacles' perspective, focusing on price takers who optimize.  
The agents are no now bith producer and consumer, and suppliers no longer  
meet the demands of traders which makes the earlier need for equilibrium in  
exchange irrelevant, and the price movement and setting is lost.  
 This system was however differentiable, and focused on efficiency (Pareto  
style).  
  
By 1918 the Vienna Colloquium, revived GE in a Walras-Cassel model,  
defining GE as a set of factor prices and output prices which set supply and  
demand equal. By using complimentary slackness, a static GE solution is  
proved to be unique in here. (A subtle shift away from Walrasian dynamics, I  
add).  
  
During the 1930's a revival of the Paretian view took place, and the use of  
calculus was beginning to make inroads. Hicks stated very clearly later his  
being influenced by Pareto, and a marginalist revival (of sorts) was taking  
place in academia.  
  
As WW II broke out, the Vienna Colloquium was disbanded, and 1939 saw the  
foundation of the Cowles foundation (with the maxim "Science is  
measurement"), and it is through the large influx of continental European  
economists who worked on Walrasian-Cassel models, into this foundation which  
was strong on Paretian Equilibrium and mathematics (going hand in hand),  
where the new consensus on what GE analysis should be.  
  
I would call this, and others have, a Neo-Walrasian General Equilibrium  
system, as it has very little to do with Walras himself - beyond banking on  
his name.  
  
So in respect to cui bono, it seems to have been a benefit for each  
succesive generation to keep working at the 'frontier' of mathematical  
economics, each proponent trying to improve on the last one's conceived  
shortfalls, ending up with nothing more than fancy algebra.  
 The ultimate 'Chinese Whisper' gone wrong.  
  
I have a brief paper on this if anyone is interested, but I have honestly  
not submitted it anywhere, and I think the gist of what I am saying is found  
above.  
  
All the best  
  
Benjamin H .Mitra-Kahn