[Question] In physical eschatology, is Aestivation a sound strategy?

In this pa­per, An­ders Sand­berg, Stu­art Arm­strong and Milan M. Cirkovic ar­gue that

If a civ­i­liza­tion wants to max­i­mize com­pu­ta­tion it ap­pears ra­tio­nal to aes­ti­vate un­til the far fu­ture in or­der to ex­ploit the low tem­per­a­ture en­vi­ron­ment: this can pro­duce a 1030 mul­ti­plier of achiev­able com­pu­ta­tion.

Later Charles H. Ben­nett, Robin Han­son, C. Jess Riedel dis­agree, claiming

In fact, while this as­sump­tion may ap­ply in the dis­tant fu­ture, our uni­verse to­day con­tains vast reser­voirs and other phys­i­cal sys­tems in non-max­i­mal en­tropy states, and com­puter-gen­er­ated en­tropy can be trans­ferred to them at the adi­a­batic con­ver­sion rate of one bit of ne­gen­tropy to erase one bit of er­ror. This can be done at any time, and is not im­proved by wait­ing for a low cos­mic back­ground tem­per­a­ture. Thus aliens need not wait to be ac­tive. As Sand­berg et al. do not provide a con­crete model of the effect they as­sert, we con­struct one and show where their in­for­mal ar­gu­ment goes wrong.

Who was right?