That first thing is a really bad idea, in a crash rather than having the upper torso getting loaded, the upper body would rotate out from under the belt. This would make the lower part of the belt the primary restraint which we know causes spinal injuries and ruptured spleens as well as other internal injuries. Not to mention the facial injuries resulting from your upper body being effectively unrestrained and impacting on other surfaces of the interior.
Raising the seat offers a fairly normal relationship of the user to the belt for a shorter user (or those with short a torso).
Yes, one could look at adapting one of the systems from another car. Hopefully there is another structural area below the existing mount. I have seen these non integral tracks bend out under load, which is fine if it was something engineered into the restraint system as part of the load reduction approach, not so good when it just occurs without intent which is what a part you or I would create. I wouldn’t try it myself on a car that doesn’t offer the feature.
The other part you linked to is reasonable, ensuring the threaded part of the mount is on the back of the plate (which is what it looks like the design is intended to do) so the threaded portion doesn’t just rip off the plate under load. The negative of it is the rotating nature of the joint which the collars shown will enable and the lack of a second mounting feature (which given this is a simple retrofit wouldn’t be expected).