A couple thoughts.
First, for main blades, any noseweight must be located well out toward the tips (not at them, though). Blades that are 10, 12 or more feet long simply aren't stiff enough in torsion for the weight to be inboard while the nose-up twisting force is mostly outboard. Bensen set the pattern for the location of noseweights; get a copy of his plans for an idea of where on the span the weights should go. Roughly 2/3-3/4 span.
Second, Chuck Beaty of course is correct that the nominal location of the blade CG to minimize twisting effects caused by misalignment of the aerodynamic center and CG of the blade is ... at the aerodynamic center. But as a footnote, at least, it's interesting to note that people have sometimes broken this rule in the past -- intentionally or not. The results have been anywhere from quite scary to mildly useful; break the rule at your own risk, though. Example:
McCutchen Skywheels are built of fiberglass, but in a pattern much like your extrusion (relatively thick walls near the trailing edge). As a result, they are tail-heavy. The amount of down-force created by the mass of the blade at the CG is a function solely of blade RPM and G-load. (Blade RPM affects it as long as the rotor is coned.) The amount of aerodynamic force at the blade's aerodynamic center, OTOH, is a function of BOTH RRPM and blade angle of attack (AOA). As a result, when blade AOA increases (initially without an increase in RRPM, which takes place over time), an imbalance of forces results that twists the blade nose-up. This, in turn, adds to the blades' initial increase in AOA. The result is a ballooning effect when you pull the stick back (or plow into an updraft). It's hard to say this effect is particularly useful, though it's harmless in small doses. On heavy gyros, it can cause control problems.
One of the early NACA reports on full-scale wind-tunnel tests of a Pitcairn gyro rotor mentions that a similar aft-CG setup was intentionally built into the test rotor. The narrative states that this arrangement is intended to twist more pitch into the blades when RRPM rises -- IOW, it was supposed to create a constant-RRPM effect. This idea appears to have been an evolutionary dead-end, though.