Image shows successful test where elements heat to red hot (~700C) without the waffle plates in place in about 2 or 3 mins without tripping the RCD as they did before the oven treatment. The item was consuming 900W, close to the stated 850W on the label.

Perhaps do not do this test for long as the elements may get hotter than they were designed for, as without the waffle plates in position to transfer heat to the thermostat, there is no working temperature control.

Comments below are an optional read.

Note: It is believed that the relatively poor electrical insulation is caused by moisture ingress changing the hygroscopic magnesium oxide to the less insulating magnesium hydroxide. The chemical reaction is reversible by applying heat.

No weight change however >0.5g was observed on oven heating. This was probably due to the amount being so small as the moisture ingress probably only occurred locally at the terminal end. (No obvious fault was observed to the ceramic sealant)

The reason why the RCD did not trip immediately, but after amount 3 mins, may be because the electrical conductivity increases with temperature as it does with many ceramics. The reason why the red hot element did not drive off the moisture was probably because it was much cooler at the terminal (see image) and there was not enough time (3 mins).

The second image shows a typical mains power consumer unit (UK). Note there are different types of trip switches. Here there are two types (i) RCD (green arrow), which is relevant to this repair and measure earth leakage and (ii) MCR (miniature circuit breaker- red arrow and are the modern equivalent of 6/16/32amp fuse wire) and not relevant here.