I agree that the final vulcanizate properties should be examined before drawing any firm conclusions and I would see whether refining the mix on a mill makes any difference. However, I suspect there is a dispersion problem here. Zinc oxide does have a reputation in this respect and it reminds us how difficult it can be to disperse sulphur in NBR.
I have encountered maldispersion of zinc oxide in my standards work although not specifically with NBR. It is important to add the activator early in the mixing cycle, with or without fillers, and certainly before the stearic acid and plasticizers or oils. Ensure the zinc oxide is dry and avoid old stocks. I agree with Metten about masterbatches: some using NBR as the binder are commercially available. Coated grades are also around. As regards particle size I am unsure about active zinc oxide - some people say it disperses more easily than normal grades, others the reverse. One advantage is that a lower level can be used.