Why more often Neutral and not Line under these conditions?

I've always had a theory that bad L connections get spotted by routine L-PE loop tests (or R1+R2) so when the reading is a bit on the high side everything gets tightened up - whereas bad N connections could go completely unnoticed (until they overheat at least).

+1 for the design of shower isolator that has twin screw terminals - and terminals fixed to the back part - so no re-arrangement of the conductors to put the cover on. (It might be better if the design aligned the screw directions (say all E-W, rather than star fashion) so that there wasn't always one that was just too tight to the wall to get a decent sized screwdriver in straight.)

   - Andy.

I note that these isolators are now being marketed as being rated at 50A rather than the traditional 45A, although I do wonder if it is simply a rebranding exercise and that the internals remain exactly the same.

There may be some substance in what you say about the cable length of the neutral when it comes to the internal wiring Graham, although this is pretty much out of our control. I do not know what the installation method details are for this latest job. The isolator is a 2 gang type mounted in a 2 gang cavity box in a stud wall. Cable looks to be 10.00mm T&E, MCB is 40A rated and sited in a dual RCD hager board which looks quite recent, although it is not labelled up in any way!

Returning to my question - is it a crime against BS7671 to omit the isolator altogether?

I note that these isolators are now being marketed as being rated at 50A rather than the traditional 45A, although I do wonder if it is simply a rebranding exercise and that the internals remain exactly the same.

There may be some substance in what you say about the cable length of the neutral when it comes to the internal wiring Graham, although this is pretty much out of our control. I do not know what the installation method details are for this latest job. The isolator is a 2 gang type mounted in a 2 gang cavity box in a stud wall. Cable looks to be 10.00mm T&E, MCB is 40A rated and sited in a dual RCD hager board which looks quite recent, although it is not labelled up in any way!

Returning to my question - is it a crime against BS7671 to omit the isolator altogether?

Get your tin hat on right now Whj , you are starting a fight here

The difference in the current standard makes very little difference between a 40-50A switches and making it =>50A reduces the operations by half when tested.
Heat rise at the terminal is a max of 45K. I think the un-operating test is 125C (don't you love BSI for mixing units in one document).
I uprated all my isolators to 50A and they passes. I also upgraded my double pole to 32A by just increasing the hole size in the terminals.