I've been watercooling since '05, so I've picked up a few things along the way. I saw that you weren't talking about using them in series, but using 2 pumps per loop. It is common knowledge in watercooling circles that the goal for flowrate is between 1.0-1.5gpm. Above that flowrate you see very marginal performance improvements, and believe it or not, it is possible to hurt performance if your pump heat dumps outweigh the increased flow performance. The performance increase in temps is plotted as a curve as flow increases, so there comes the point above 1.5gpm that your PQ curve starts to dramatically flatten out reflecting that as flow increases above 1.5gpm, the resulting decrease in temp lessens. With the loops you are mentioning, I see no reason to run dual pumps and seriously doubt that you would see any measurable temp decrease by running dual pumps over a single per loop. I'm not sure what CFD analysis you are doing, but you have to look at PQ curves for the given blocks and rads to see where you would be at. I would recommend taking a look around Xtremesystems watercooling forum for more information. There is also a pretty good flow estimator linked to there in the sticky (http://www.xtremesystems.org/forums/showthread.php?206542-Water-Cooling-Estimator-Spreadsheet) that is pretty much plug and play. So does your new job have the kind of tooling setup your old one did ie CNC, etc.?