Instead of deorbiting the second stage (particularly with Falcon Heavy) can they be sent into a common parking orbit, there must be a lot of useful materials, not to mention engines, that can be used in the construction of space structure. Modern fabrication techniques would aid here, 3D print etc. The second stage could be boosted into a more stable orbit if it could be partially refuelled.
That does not make much sense, unfortunately: Most payloads are launched into a Geostationary Transfer Orbit. Such an orbit is highly elliptic. If the second stage could reach the Geostationary Orbit, it would just go there directly and remain there with the payload.
So in order to reuse the second stages you would need to cluster (ideally dock) them in roughly the same spot on a highly elliptical but stable orbit. This will cost you money since parking (and carrying docking capabilities) requires more fuel. Hitting that spot constrains your launch window severely. Unfortunately your launch window for a geostationary satellite is already quite constrained (you want your apogee above your target, if you wait too long before raising your perigee your orbit will decay). So your orbital mechanics would be more complicated.
Finally, in order to do something with the spent stages, you would have to climb up to your parking orbit (i.e. a close-to GTO). This requires a lot of fuel and basically halves your payload compared to the Low Earth Orbit. As long as you do not need many used stages, you are better off launching your new engines into the LEO.
tl;dr: Too complicated, no real benefit
There is a parking orbit that GTO launches often send their upper stages, that avoids the GEO belt.
However it is possible that you underestimate how big space is and how far apart everything is.