Where did you get the idea that changing the logo / model number has a significant effect on performance ? There are basically 3 different categories of cards:
a) Reference cards ... shown in the chart. There is no significant difference in performance between brands and the performance of EVERY 780 reference card is indicated in the chart at as 84%.
b) Non Reference cards ... the chart shows BOTH the performance of the reference card at 97% and that of an MSI AIB Gaming X card at 3% faster. Having built several boxes with 970s AIB cards i can tell you from personal experience that there is little significant difference in performance. While they may come with slightly different overclocks, manual overclocking puts the pretty close to the same level.
c) the super cards ;ike the MSI Lifgtning, EVGA C;lassified but since they have become irrelevant and are not within the discussion, we can drop those.
Here we see a better model than yours from Gigabyte reaching an overclock of 1516.
http://www.guru3d.com/articles_pages/gigabyte_geforce_gtx_970_g1_gaming_review,26.html
Here we see the MSI reaching an overclock of 1501
http://www.guru3d.com/articles_pages/msi_geforce_gtx_970_gaming_review,26.html
The above clearly shows the changing the logo or model number of cards changes little with respect to comparing different generation cards. Yes I would rather have 1516 than 1501.... but you must recognize that this is less than a 1% performance difference.
The graph DOES SHOW you quite clearly that the difference between the stock card and the AIB card is 3%. The above shows you that the competition is extremely close, the two different model cards vary by less than 1 % and worse case it might be 2 -3 %. Again significant if deciding between two current generation cards but cross generation.
The graph
clearly and factually answers you question... the AIB 960 card is 16% faster than the stock (reference) 780 card. No model number or logo change will change that answer to any significant degree ...maybe it will be 13% ... maybe it will be 19% but in no case no matter how many non-applicable variables you throw at it, the answer will still and always be that the 970 is faster.
I should note that I have two 780s (water cooled w/ 22% OC) and two 970s (19% OC on air) in this room and the benchmarks used include overclocking. Believe what you want, but as you can already see, you won't get a different answer.