On Mon, 27 Nov 2017 09:02:08 -0800 (PST), whisky-dave
wrote:

snip dave throwing his toys out of the pram as it's obvious he's dug
himself in too deep he now can't get back out

'Yes dave, you are right and I am wrong.'

Right, if it was me, I would do the following tests, just OOI.

Switch the NoII / 1300W position on *only* with your current
measurement plug in place, and the main stat on full, see if it cuts
back to 700W or *at all* or after the ~2 hours you mentioned
previously.

If it still overtemps with just the high power element on, try it
again with just the low power element (NoI, 700W, it really shouldn't
with just that because 1) you said it doesn't anyway 2) the rad should
be able to dissipate 700W (when not covered)).

Cheers, T i m