One more picture; result of pull test. The line broke in halfway between wrapped ends. I don't know the break strenght of 19-strand 0.015" cable, I'll guess it is less than 20kp.
Also, I think you exaggerate the effect that soldering has to temper of cable. Even the most sensitive tool steels are relieved to spring temper at about 300C, I doubt that any cable used in our toys is left that hard. Bigger risks are the remains of acid flux and possible stress risers.
I agree that a normal soldering operation shouldn't get hot enough to seriously anneal stainless cable. Our cable is typically 300 series stainless and it doesn't really harden except work hardening during the wire draw and spooling process.
I agree that the biggest threst is from the flux that can be trapped inside the cable under the solder that cannot be neutralized or removed. It then corrodes the wire and seriously weakens it.
I do not agree with soldering steel cables used in stressed conditions.
I think your method with kevlar thread is likely to be quite secure. It does seem to require a lot of work as compared to simple dimpling of a copper tube however.
I assume you cover the thread with CA and opaque shrink tubing to reduce the effect of UV light.