OK, I'll take the bait

One company named Comp-Air has already been building big, square, composite high wing bushplane-type airplanes for a few years. I have no idea how good or bad their airplane actually is, but for this discussion let's say that it represents a reasonable demonstration of a composite version of the traditional conventional layout.
You said that you wanted somewhat conventional layout, so a larger version of Barnaby's Facetmobile (which reportedly had surprisingly good STOL ability) is not relevant to this exercise. Same goes for giant flying wings, jump-capable gyrocopters, Burnelli lifting bodies, and tandem-wing Rutan Grizzly thingies.
A plastic version of the Porter or Gippsland Airvan is pretty close to what the Comp-Air is trying to do. A plastic copy of the PA-14 or stretched PA-22/Bushmaster is probably too obvious,and both the Glastar and the smallest Comp-Air gets into that range already.
Out of what's left, I'd suggest looking at a scaled down Fairchild C-123 or Short Skyvan layout, executed in composites. The reasons for this are fairy obvious:
- Best prop clearance, making use of high wing
- Best visibility in all directions
- Best cabin volume vs. overall aircraft size footprint
- Better span loading / spar bending moment
- Fuselage able to be built using flat or nearly flat sandwich panels
- Stand-up and walk-through cabin
- Possibility of rear loading ramp instead of doors
- Can retract wheels into fuselage blisters (C-123/C-130 style) resulting in DRAG and RANGE improvements
- Tailwheel config. is optional, but not mandatory (already have big prop clearance)
Of course the big price to be paid for this is that is needs two smaller engines instead of one big one. But, even this disadvantage is offset by additional advantages:
- Twin engines provide flight-continuation capability over desolate terrain
- More speed vs. economy choices to suit different missions
- Better propeller/thrust efficiency (no bulky fuselage affecting outflow)
- Possible use of augmented lift from propellers if designed right
Using the successes of Wittman's Tailwind and perhaps a few others (Taylorcraft), it can be shown that flat sides and square corners can go reasonably fast. If you can build the fuselage with flat sides, then you can do without enormous, complex molds for the fuselage, saving you a tremendous amount of money. You might be able to laminate and bag your sandwich skins on a flat, or single curvature table. Michel Colomban has achieved exceptional performance with the MC-30 ultralight, building the fuselage as single-curvatures using simple mold/support tables. In this application, your benefit is not so much the faster speed, but that the drag reduction will allow greater fuel range, which is important in remote operations. This is also the justification for building out the blisters on the lower fuselage to retract the wheels into... not the speed but the efficiency to to have big tires when you need them but keeping them faired in when you don't.
Also, this airplane might be a starting point, if you happen to have a large tubular mold available...
http://en.wikipedia.org/wiki/IAI_Arava