OK. Well, for a start your chances of getting this to 2000 (C or F) are probably slim unless you poked an awful lot of power into the system Which, btw I am not sure I fully understand. One reason for this- and there may be others entirely unrelated - is that it has huge radiative surface area. Somehow I think your salt-water caps would boil dry before it gets too hot. But lets assume that it does. The 'stone granite compound' will almost certainly start to crepitate long before that - possibly with the ejection
https://en.oxforddictionaries.com/definition/crepitate of stone chips caused by the contrast between the low thermal conductivity of stone materials and the rapid heating you envisage.
Pulse power right down to low frequencies is not too difficult to organise. Would that be the answer? If so we can talk about that some more.