Canon's 35L doesn't seem to fare much better according to Photozone, can't say for myself since I haven't used 35L.
When it comes to bokeh, I find PZ to be pretty useless. I do not find their sharpness charts much useful either.
I own the 35L. Tell me what you want me to prove to you - that it has great bokeh, or that it has poor one. I can prove both with examples.
Well, if I really wanted to characterize an objective, I'd put it in the MTF bench at work. Or start counting wavefront fringes in a double pass interferometer setup. Or check the actual lens performance from construction data. But online sites are far more convenient for referencing and save me from quite a bit of hassle.
So I don't have a great interest in finding out which one of the 35s has the best or worst bokeh in which situation. The point was that the difference between Sigma's and Canon's are comparatively small, and in some situations the other is better than the other and vice versa.