This is the idea of a screw or bolt with some new features that might improve its performance.
It has an external and internal torx head. The best possible shape of the exterior part of the head in order to keep an approximately constant thickness while having an internal torx head, is just an external torx head.

The section of the screw thread closest to the head has to bear more load than the tip, where there are just a few threads pulling the screw. Therefore, the load-carrying section of the screw can be much smaller near the tip than near the head, and a screw with a conical longitudinal hole can support the same load while being lighter. Also, the internal tube could provide some grabbing feature similar to that of Robertson screws, but only for screwdrivers that intend to do so, with a tip similar to TTAP. The grabbing feature is very useful when working by hand, but it's detrimental for machine use. These screws can be used in both situations, allowing reusability of screws in mass-produced items.

Variable pitch for long bolts with a known load:
When a longitudinal stress is applied to the head of the screw, the first few threads will deform slightly to support that load, reducing the remaining load transferred to the following threads, which will deform less, supporting less stress, and so on until the last thread, which supports almost no load. To distribute the load more equally between the threads, I thought about designing the screw with a slightly smaller pitch near the head, that, after applying a tensile load and causing a strain, will deform to the right constant pitch. When the load is applied, the threads near the tip will be the first ones to support the load, then the screw will stretch, and the following ones will come in contact and support the load. As the load increases the screw stretches, and more threads come into contact with the threaded hole, improving the load distribution. This could allow the use of screws with a smaller diameter when the materials present are not very resistant. Notice that this modification would be detrimental in compression, and the pitch should vary in the opposite way. A flexible core could maybe have an equivalent effect in tension and compression.

I have not been able to adjust the angle of the threads in the swept cut, so the triangular threads just transform into trapezoidal ones. When I get what I want I will update it. Maybe a normal screw design could be compressed longitudinally to achi...