Another factor against going from balsa wing skins to composite, besides the thickness of the balsa as Tom noted, is that the foam cores are low density "white foam". If you want to to with a composite wing using fiberglass/graphite skins, the white foam would be too fragile.
Also BTW, white foam cores like those are meant to be bagged at around 5 to 7" hg of vacuum. If you used a strong vacuum as usually used for fiberglass/graphite or other composite layups, like 15 to 20", you will CRUSH the white foam. I onetime accidentally bagged a left wing of a white core at about 10" vacuum, and that wing's core was permanently compressed, it ended up about 1/16" thinner at the root than the right wing (7/16" vs 1/2"). The model was flyable but not vey good. And the left wing stalled before the right wing did.
So, if you do the original white core wing with balsa skins, be careful not to set the vacuum level too high or the white foam will crush.
If you really truly wanted to do a composite wing, then you'd need to get new cores cut, for the correct airfoil thickness and using blue foam of some other stiffer foam ("Spyderfoam" is the best but a bit pricey). I get cores cut by "Flying Foam".
But I really think the model is better off with the original balsa and white foam core method. I will say that since I like slower boosting gliders, that if i was building a Phoenix, I'd replace the medium to heavy-ish balsa skins in the kit with the lightest balsa I could get, and boost on G12 power. I would add a bit of glass cloth under the skin near the center section and perhaps a thin graphite strip mostly spanwise under the skin as a "spar". I do not recall for sure if the Phoenix kit used some 3/4" or whatever width of fiberglass tape between skin and core as a sort of spar.
I do know that the regular model was designed to hold up to a lot of power, IIRC has flown and held up to G40 power. But anything beyond that might shred it. I don't like crazy-fast boosts though, and the model has to be built so heavy to take such a boost velocity that I'd rather go up slower on boost and have a much lighter model for the glide.
Also, if I built a Phoenix, I do not think I'd bother with the rudder, just to save weight. Even though I mainly fly 2-channel models with rudder-elevator (or 3 ch with throttle) and do not fly aileron models much. The good thing is that the Phoenix has some dihedral and handles nicely with ailerons. I've never owned one, but got to be the test pilot for a couple of newly built models that some college students were using for a rocket oriented "Ultimate Egg Drop" competition. The dihedral (plus sweepback dihedral effect) made them easier to fly than I had expected since I had very little aileron model experience at the time (enough not to be a danger, but not enough to feel very confident. But in a not-ideal situation that was the best option).
FWIW - heres a page on what I did to modify an old Vector Aero Cuda rocket glider to use a center flap to try to land more accurately. I also totally revised the fuselage layout. But that it not the point it is to explain why its so different.
The point is how, if you used lighter balsa for the skins, you could do a thin graphite spar strip under the skins, as well as some fiberglass cloth as needed. for this one I had to add some fiberglass cloth to make of for the loss of structural integrity with the cut-out flap.
Also one of my biggest tips for spreading epoxy for an even thin amount is to use a bit of dye pigment, I used pigment powder, to make the epoxy red. So when I squeeged the epoxy out into a thin sheen of epoxy, the wood had a light pinkish appearance. Any areas with too much epoxy looked darker, in more need of squeegeeing to remove the excess. This also made sure there were no missed areas. Very hard to tell that when the epoxy is clear and not dyed. FWIW, anyway. I do not do that dye pigment with composite wing layups though, other than for the leading edges.
https://georgesrockets.com/GRP/GLIDERS/CUDA/CudaFlap.htm