Ok, being the numbers geek I am and having spent all afternoon being deluded with emails and messages asking if this is possible, I sat down with a spreadsheet and did the hardcore indepth numbers.

The entire potential of this working depends on two factors.

a) That Delta can keep TOTAL costs constant, even when increasing capacity.
b) That the relative yield dilution factor does not exceed a moving threshold of relative capacity increase (eg. On 10% systemwide capacity increase, the dilution must not exceed 5.8%)

Neither of the two is particularly easy to achieve, and of course, as the costs increase in (a), the threshold moves in the wrong direction in (b).

My analysis and opinion is that this does not make sense in the form that has been suggested, but a modified variant where larger aircraft are diverted to lower yield hub-bypass flying in markets which can support that capacity would definitely be something worth considering. My sources at Delta seem to point towards that conclusion as well.