Removing enamel with molten alkali

The details for removing enamel with molten lye are frequently kept hidden on the grounds that the technique is dangerous. My own assessment is that it is comparably dangerous to leaving work on Oxford Street in London on winter evenings, given suitable safety training for both activities. Other people's assessment may differ.
I cooked the object with the enamelled areas covered with about an equal volume of anhydrous sodium hydroxide ('caustic soda' or 'lye'), at about 400ºC (which is above the melting point of about 360ºC), in a tiny iron container (an old pot lid) in a small enamel kiln. I imagine a tiny copper pot would work perfectly well. The molten lye dissolved the enamel within minutes, and was then allowed to cool when it set rather rapidly into a grey-green mass. When sufficiently cold I irrigated it with a large amount (a bucketful) of cold then boiling water to dissolve the caustic soda together with reaction products. Usually I found that the enamelled object would be stuck to the pot with solid lye underneath, hence the boiling water to help dissolve it quickly.
See the footnote to this website for the disclaimer concerning safety! Lye is very caustic to skin (and in fact all human tissue) even at low concentrations and temperatures, so I wore goggles and rubber gloves. Solid cold lye can misbehave with water since it is an exothermic reaction, and I found that molten lye misbehaves even worse (explosively possibly) with even very small quantities of lye - hence the reason for waiting until all reaction products are cold, to avoid the risk of splattering hot caustic alkali around. Good ventilation is required - a small amount of injurious alkaline spray is produced which one wouldn't want to breathe in, and more if the alkali is not cold when added to water.
But, the pay off - the vast majority or all of the enamel simply dissolved away, and in those cases where some remained, it was removed by drying carefully then repeating the operation.
An alternative technique uses an aqueous paste of sodium chloride (common table salt) and potassium sodium tartrate (Rochelle Salt), applied to the enamel areas, then the whole heated to 750ºC. It is then plunged into cold water ('ice cold' is recommended on various sites). I found this was occasionally very successful, but sometimes only as good as plunging red hot enamel into cold water, and that it often never removed all the enamel even after repeated application. The use of such high temperatures and high temperature gradients on quenching is also likely to be problematic in some cases, causing warping for example. The chemistry involved seems to be unknown - the enamel is not dissolved, but seems to break away from the substrate more easily.
I also briefly experimented with molten potassium hydroxide for removing enamel with satisfactory results. The melting point is similar to that for the sodium salt; and the boiling points, at above 1300ºC for both, are sufficiently high to guarantee a stable temperature region of molten alkali. The potassium salt is of course potentially more reactive, and poses greater safety risks. It didn't seem to improve the removal of enamel, but was more effective than the sodium salt at removing ceramic shell investment from an old bronze specimen I had had lying around for a few years. As expected, it also severely degrades glass.