I think I owe some actual informations about cycle life tests of Samsung 21700 cells. At first I must repeat that I can not show you graphs but I can tell you numbers.

At this time I already reched milestone of 1000 cycles with INR21700-50E under several conditions. I also have 1000 cycles with INR21700-40T under one particular test 0.5C-1C at 100% DoD and another test conditions are already running.
I personally use dividing cells into three categories:
HE (High Energy) - LG MJ1,M36,M50, Samsung 35E,50E, Sony VC7, Sanyo GA...
"BP" (Balanced Parameters) - LG HG2, Samsung 30Q, 40T, Sony VTC6...
HD/HP (High Drain/Power) - Samsung 30T, 25S...
I am not focusing on HD cells, so I have only limited database/knownledge of their cycle life test. But I have tested all above mentioned cells and many others from HE and BP categories in small cylinder format. Here are some general notes
1) Absolute best cycle life is domain of HE cells!!! but the continous discharge rate must be lower than 1C and charge rate 0.5C. I personally recomend to design traction battery with HE cells up to 0.5C average discharge and up to 0.3C average charge rate. HE cells at higher continuous discharge rates than 1C are going hot as hell and their cycle life falls rapidly to ca 150 cycles at 3C continuous discharge, also 1C continuous charge rate usually have significant negative impact on their cycle life.
2) Almost all tested BP cells have significantly worse cycle life than HE cells. But their cycle life is usually almost the same under 1C or even 3C continuous discharge rate. Also their cycle life in not affected if they are charged at 0.5C or 1C continuous rate. Also BP cells are much cooler at higher continuous charge/discharge rates.
Some data for Samsung cells:
50E: holds 83% of its nominal capacity after 1000 cycles at 0.5C-1C 100% DoD
40T: holds 63% of its nominal capacity after 1000 cycles at 0.5C-1C 100% DoD
40T: holds 77% of its nominal capacity after 350 cycles at 0.5C-3C 100% DoD (it is the same number as for 0.5C-1C at 350cycles)