@elsparko
I would stop this practice asap, regardless of who taught you this or what field of electrics they worked in, this is poor to bad practice, I would also question the quality of their working practices given they felt this was an acceptable method.
The glands are manufactured to certain tolerances and the SWA has to also be to similar tolerances so that the glands are more or less universal to the given size on their packets thus will fit what it says.
The glands rely on a full use of the strands to maintain an even grip all round, taking a number of strands away and routing them direct to the earth will give a poorer glanding termination and the remaining steel strands you have secured will not satisfy the reg's, under fault conditions the gland or strands brought through could fail and/or see damage done to the cable.
Your R1 + R2 tests and ELI do not show all issues and when 1000+ amps flow under fault conditions your gland should be fitted as per manufacturers instruction as the regs require you do.
If you secure and tighten your gland correctly it should not come loose, if due to environmental conditions you suspect the glands could loosen then serrated washes should be fitted, there is also the case that the installation requires regular inspection and this should pull up on any glanding issues.
In my experience of 30+ years I have never seen a gland come loose that was fitted correctly and suitably done to account for environmental conditions so I do not see this practice serves any purpose, the only glands I find loose etc are down to poor termination practices of which I would put your method as one.