Isn't there a pretty easy way to settle this that will remove unknown factors?
1- Create two different receiving addresses for a wallet in a neoscrypt coin.
2- In a system with two equivalent cards run one instance of hsrminer and one instance of klaust ccminer towards the same pool but use one wallet address for each process.
3- Let it run for an hour
4- Inspect earnings as reported by pool.
This would make both systems run on the same difficulty levels at the same time and therefore remove as many differences as possible. Of course luck will enter into the picture, but I imagine that one hour should be enough to let this even out. Otherwise it's always possible to run longer to rule out luck as a factor (not completely of course but to a reasonable level).
Yes, that would be the next approach I would use. I was already running a comparison on my two separate GTX 1080 systems which is why I did a consecutive 24 hour test of hsrminer vs. ccminer. Currently those two systems are doing a head-to-head comparison of dstm vs. bminer mining ZEN on luckpool.org and I was next going to do a comparison between zergpool and hashrefinery because preliminary testing of both has proved very promising.
Yes I will try to do a test as well during the weekend to be able to get a personal verdict.
I have no access to several rigs, but I have several cards of the same type so I should be able to set it up as described fairly quickly.
What is promising?