Ok dude.. One day you'll mine 400 coins, the next day you'll mine 40 coins. You can look at whattomine or any other website and they will give you the snapshot of how your coin forecast will be for that exact second. If you hit refresh in 5 mins, it can give you a number 10x that or half that. Concurrent mining is a harder way to do it but it would need to be the same pool, same coin. The simple way is just measuring hash rate and shares submitted at the pool for each miner over a 24hr period. I've already done it, It's 8% higher on hsrminer, I'm ok with that given the 1% fee.. I don't think it's anywhere near 8% higher with the non Ti cards, so it's very possible hsrminer isn't worth it for those but I don't have non Ti cards so I can't confirm..
Yeah, you keep talking about getting the *current* difficulty from whattomine while I keep saying I am using the *average difficulty for 24h* from minethecoin; very different, those two.
Now, there is one potential problem with using the average difficulty from minethecoin and applying it to the pool: minethecoin is likely performing a moving average for all of the blocks found by all of the pools for the past 24 hours, whereas any given pool will only find a fraction of the total blocks found in a 24 hour period. But I anticipated this issue and it is one of the reasons why I chose the official Trezarcoin pool to do my test and ran the test for 24 hours - again, that allows me to use the average difficulty value from minethecoin and it helps reduce the effects of wild variations in pool luck and/or of the difficulty of the blocks the pool does find.
I already agreed that running the test concurrently with identical hardware on the same pool, just different wallet addresses, would be a better test, but I don't think that automatically means my methodology is worthless.