Your test wasn't controlled, therefore, your results mean - nothing -.
Sorry man, that's just the way it is. The entire reason we control the variables in our tests and take the time to do them correctly is so we can actually trust the results. If you were really sick and the doctor told you specifically that he wanted to do a 12 hour fasting blood test on you, how accurate would the test be if you ignored his direction, or if he/she didn't keep track of the vials, or the lab tech was in a rush and didn't do his job correctly?
A quick 2 minutes in the camera store test without controls won't tell you much at all.
My old saying "nothing could be more potentially misleading than a poorly done, or incompletely done, test" stands true.
So, what you should have done (at a minimum) is to make SURE focus on both lenses was absolutely in the same spot and perfect. We have to realize that even slight misfocus can significantly remove sharpness. A tripod to remove motion blur would have helped too. The more you sit back and think about how many things can conspire against you to remove resolution, the more you'll see why I'm a bit "harsh" in my comments about controlling the test properly. (high ISO is another, subject movement, etc, etc is another) You could compare an Otus and a kit lens in the manner you did and get the same kind of results. The reality is that every objective test, and most every subjective test done on those two lenses says the Sigma art is sharper, particularly in the wider apertures. So if you aren't getting results that approach the consensus, that should be one big red flag that you didn't test properly.
-m
I'm a wedding and portrait photographer. I tested both lenses for over an hour and have a multitude of examples. I wanted to test the lenses color fringing under high contrast subjects, I tested under backlighting, side lighting , sharpness wide open, stopped down, vignetting, distortion, etc etc etc. I simply don't own a laboratory and, in my humble experience, people will always challenge the testing methods and environments.
I'll be 100% honest with you. I could care less about testing them in a lab. I'm a wedding and portrait photographer and so I tested them under conditions that made the most sense to me (I don't shoot test charts for a living). The results were able to be duplicated by myself on several camera bodies. If I was at a wedding or a portrait photo-shoot, I'm confident that I could duplicate the results. This was not a situation where "sometimes" the Nikon lens was better......it was identical or better in 99% of the images I took.
While I respect your opinion, I do not know of ANYONE that has tested these two lenses. People have claimed to have done so, or have owned them and different time periods.....but to the best of my knowledge, I am the only one that has RAW files of both lenses being tested on the same camera, pointed at the same subject, under the same lighting, at the same time. Thus, my review is far more revealing than all the conjecture on the internet.
The point is that the Sigma 135mm ART lens is rated very highly and people love the lens. Everything I've heard, read or seen has been nothing but assumptions that the Sigma 135mm was "better" than the Nikon 135mm. But I have yet to see any type of controlled test (or any test for that matter), I have yet to see anyone compare RAW files........I could go on, but I'll assume you're intelligent and understand my point.
Could it be the Sigma 135mm lens I tested was a poor copy? Absolutely.
Could it be that the Nikon 135mm was the best copy in existence? Absolutely.
Could it also be that the Nikon 135mm is closer in performance to the Sigma 135mm than people want to admit? Absolutely.
Regardless, what I am trying to point out is that there should be further testing done between these lenses (with RAW files to prove the results) to see if my findings are accurate and can be duplicated by others. From my findings, it should not be assumed that the Sigma 135mm is a better lens. But I simply do not know of any other source that has tested both of these lenses.