There are several considerations in screen layout, and with 100+ hours on a dual 10" screen Garmin G3X touch installation, these ideas (biases? prejudices?) have some experience behind them:
* Multiple 7" screens will waste panel real estate with the bezels. A big screen won't do this;
* If you need to reference the two screens simultaneously, the big bezels are a visual barrier that makes the transition harder. The old Socata Trinidad has an overly styled panel that was clunky to read because the styling put big visual barriers between sections of the panel;
* The 10" screen is big enough for a column of engine instruments plus PFD and MFD. And if you want additional engine instruments, or larger scale, that's an easy touch to get it displayed or to shrink it;
* Moving map is really a hazard display, more than it is a navigational instrument. The hazards it shows are weather, terrain, traffic, and airspaces. You don't need all these hazard displayed all the time on every flight, although there will be times that you will want them up continuously for at least a short period of time;
* Wider PFDs are easier to read the attitude on than narrower displays because at some point your eyes start picking up peripheral cues;
* Most engine instrument displays on glass displays are *terrible*. No tick marks on the power displays, and that doesn't help if, say, on an ILS you want to reduce power by 1/2". You end up doing a lot of math instead of flying the airplane. Some engine status instruments (temps and pressures) are little inconspicuous sliders on brightly colored backgrounds. Some vendors' display are worse than others. You do get used to it, or maybe you just get used to the extra time it takes to read the displays. In any case, if you've got alerts and warnings programmed in, that helps compensate for poor status displays;
* On my RV-9A, each screen has engine instruments outboard, then PFD, then MFD in the center. This means that each pilot in effect has two MFDs;
* Because I bought the plane already flying, I couldn't do as much as I wanted to with the panel. Specifically, I wanted the two 10" screens as close to the GTN 650 as I could get them. I've learned that it's hard to read a full screen display on the far side of the cockpit, but it's not bad reading the near half of a split screen display;
* Stein suggests that if you're going IFR and typing in a bunch of waypoints, get the GTN 750 instead of the GTN 650. It also puts another usable MFD screen equivalent in play;
* It's nice to have a standby attitude display, but I'd also like to have a standby GPS panel mounted so in case you end up flying with battery powered instruments, it's all on the panel -- moving your head up and down to look at an iPad on your lap is an invitation to really bad things. You can real quick run out of panel space, though...
It's tempting to say that lots of this doesn't matter, that you get used to whatever you decide on. However... back in the early days of the Macintosh computer, they ran tests to see whether arrow keys or a cursor with a mouse was faster. Lots of folks, used to arrow keys, swore that the arrow keys were faster. They were, in fact, slower. What happened was that there was a kind of amnesia and people didn't pay attention to time spent fussing with the arrow keys, so the arrow keys were left off. (But arrow keys did make a comeback...) There are numerous similar things that can happen in avionics, such as slewing a cursor with a velocity control joystick. Another factor that can pop up is that there is a context switch between manipulating the data directly and controlling the display -- such as switching screens. That context switch is non-productive workload, and can add to accumulated stress and fatigue on a challenging flight or on a bad day.
I do envy the RV-10 and RV-14 guys because my RV-9A panel is not quite big enough for everything, plus it's not feasible to change the panel. Maybe some day I'll trade the -8 and the -9A for a nice -10...