You were only criticised because you were wrong.
Don't expect BAS to accept that, even though it is correct.
I had missed that TTC had also suggested it by the time I saw BAS suggesting it - and replied to BAS's later post. So it wasn't a case of having a go at BAS and ignoring TTC - it was simply a case of replying to the last person to suggest it.
IMO there is only one* possible reason to have a "cut the power to turn the light ON" switch, and that's if you are somehow required to provide that function and cannot avoid having only a 2C+E cable between switch and light.
Otherwise you simply loop at the light (only needs 2C+E supply, and 2C+E to switch), or loop at the switch (which requires a 2C+E from switch to light), and have the switch provide switched power to the switched line input on the light. Done that way, there are no "surprises", nothing for the user to learn from when they forget to switch it off and it damages the battery, it just works as intended - switch on, light on; switch off light off; power fail, light on till power restored or battery runs down.
So given that the requirement above is going to be "quite rare" (and certainly not applicable in this case as the switch is going to be near the light), a general suggestion to use that setup would fall foul of "good design" and "good workmanship" general requirements in the wiring regs. As such, it would not meet wiring regs, hence is wrong- unless you can argue that deliberately designing a circuit in a very suboptimal way is good design
* I don't count a test switch in this as that's there for testing, not for functional switching. And if you also argue that "the light doesn't have a maintained function" is a reason then I'd counter that it's against the "good selection of materials" general requirement in the wiring regs.