As long as you use lipofection, the co-transfection would not likely to produce a cell with one of each vector transfection. When you use multiple plasmids, each cells within a population may have different expression level or intake of the plasmids amount, but the RATIO being transfected into a single cell should remain the same. That is because even if you added a very small amount of plasmids ( for exampe 100:1 ratio), as log as it is ug or ng order, there is a number of each plasmid copies in some molecular level, and they are transferred into a cell as a group. If you want to see this you can co-transfect GFP and RFP and do FACS analysis. But, most people do cotrasfection and IP or luciferase reporter assay with dual color system. I have never heard any problem with those study, so essentially it will not happen.
Anyway, using cas9-2A-GFP or -puro is very easy and excellent way, but I experienced that for some cells with low transfection efficiency, GFP or maybe Cas9 expression seems to become inefficient. I suppose this is because the larger plasmid or protein tends to be transcribed less efficiently. So, for some types of cells it may be better to co-transfect rather than single all-in one vector or -2A system.
Now, I would like to know what others experienced or any suggestion.
SF