While I generally welcome the possibiliy to change the tessellation factor, the AMD optimized setting has to go, especially at default - no matter if profiles exist or not. If they were to exist, it would be cheating, if they don't as it is now it is a useless option. I strongly suspect AMD implemented this with cheating in mind but when they were able to improve on their tessellation, they simply abandoned it.
As for 3Dmark11 and benchmarking in general I don't really see the reason for modifying the workload. Why not disable other things while they're at it? Where do you draw the line? Just my 2 cents.
Do you bench ? i think we use absolutely all is possible to do for get a score.. disable mipmap, put AF off, put AA off...
Nvidia have even some harder setting who let you use bilinear filtering instead of trilinear, play on LOD etc etc .. ( when AMD have not ). I have start bench with the 9700Pro ( Maya edition ), then i have use 2x 6600GT in SLI and 6800, i have then use a X1900XTX+ a X1950XTX in CFX and continue years after years to use Nvidia or AMD for do it ... For bench i have not a preference, i like use what can give some good score.. I have even use 2x 2900XTX for bench in Crossfire, as this setup was absolutely unbeatable by Nvidia cards. ( for bench, so without AA ofc and with high OC ).
Im not completely right with it, personally i will have vote on Hwbot for dont let this setting be used ... but now i have no choice, cause if i dont use it, my score will look miserable against other 7970's . ( who is not so much important for me now, as i have stop since a while to bench ( anyway on extreme level, i still like release some little score time to time, if you are not pay for do it, benching cost really too much money now .. )
As for Tesselation setting.. as you know, it is absolutely not effective in games..
I have never see anyone claim yet the tesselation level is used in profile..
Dont forget this is ATI who have invented the tesselation, at start this was a technic for increase the number of triangle on the frame without decreasing the performance, so ATI/AMD with Miscrosoft have fix a maximum level for it .... AMD ATI have keep this level for DirectX11 ... they have not try to get higher of it, it was normally not needed.
When you have 80%+ of the workload of a gpu who is dedicated for "render" tesselation, this is absolutely not why this technic have been designed for. Specially when this is for render planar surface who dont need to be tesselated.