Don't set your mining GPUs as primary, simple.

Great!  Like I said, I have asked here and in the 1060 thread if that works because I don't have a 3GB card to try, but already suggested that was the fix.  As long as it's not pci(1:0:0) then it should be fine, which is what you're saying, right?  My 3GB 1060 is off for repairs or I would try it myself.  If you have imbedded graphics in your mobo, it shouldn't be a problem, either.

0.17$?? wow, is it still profitable for you to mine? I pay around 0.06$

that's insane..come to Ukraine

Ikr... Well I knew it from the beginning that power cost would take half of my profits... So I guess I have learned to live with it. I got my ROI early so I got that going for me, which is nice  
"Greeks pay your taxes" they say... well we do...   (80% of that 0.3$ is taxes)

I doubt it makes any difference, it's just a matter of convinience. The config file is parsed and passes the parameters to the miner at runtime anyway. If you change pools or use different settings often, a .bat file for each pool is more convininet than having to adjust the config file.

The same parameters work for both the .bat or config files, so as long as the parameters are set correctly, it should make no difference which one you used. If you look at the log file, the first line is the parameters passed to the miner. You can also use different config files in the same way you use different .bat files. You just need to put the name of the config file to use when you start the miner.

And i was hoping you would advise they guy to set us up as primary pool 

1060 w/3GB RAM should do about that, maybe 21Mh/s if you get things just right.  My Gigabyte card likes 70% power limit, about +220 GPU speed offset and +1000 memory offset.  Not all cards will do that, I have Hynix VRAM, also.  Some might get a little more, but I don't know of any way to set the memory more than +1000 offset.  My card will hold a stable +240 GPU, but I get crappy shares relative to what I get at +220.  Every single card is just a little different unless they came out of the same batch, same model from the same maker.  I have my temp max at 80C which is about what it does when I'm dual mining (and yes, about 350Mh/s for Verge XVG is what I get, although you can tweak -dcri up to the best your card will do without impacting your ETH hashing much (I limit it to 1% impact) then you might be able to get 400+Mh/s but in my experience it messes things up and I get recycles way too often instead of it running the clock out (-r 239) at four hours, which is my self-imposed limit before I restart the miner.  I was using -dcri 70 which seemed to be the sweet spot for intensity on my specific card.  The other rig is a GTX1080 with RX 580's, all 8GB mostly Asus OC/ROG/Strix models with one Sapphire Nitro+ 580 OC.

Some people insist on just letting it run forever, but my machines seem to grow a minds of their own after a few hours, so 4 hours is what I picked.  For the GTX1060 memory ramped up to 10K speed, I build the DAG in between 8 and 9 seconds.  No big woop...  I am back to single mining ETH for now, I just get too darned squirrelly dual mining.

You can not change the nVidia GPU BIOS for GTX10x0 cards, it's encrypted and won't boot if you mess with it (which is assuming you are truly a crypto genius and can figure out the encryption, which nobody else has - so far, anyway), so you are stuck with mods using AfterBurner or nVidiaInspector, etc because Claymore doesn't do memory tweaks for nVidia (or fans, temps or power limits) like it does for AMD cards.