Your second scenario is about the only circumstance I can see this being useful. The large, sparsely populated western states with a lot of towns too small to ever be able to support a station of their own and rely on translators from a larger market. This would give businesses in those smaller towns the ability to advertise without paying big-city rates.There are only a few specific situations I could foresee zonecasting being useful:
1) Urban cores that are covered by a transmitter 10+ miles out (thinking of downtown Seattle or Everett, WA) where the main signal gets hit hard with multipath.
2) Places such as the Wasatch Front and their relatively recent move-ins that are physically terrain shielded from their target market (these stations already have non-zonecast boosters in place…and frequently have issues keeping the main transmitter on)
The logistics of this could be daunting though. While most automation systems allow for "slave" systems, you'd still need a workstation for each signal that you tend to utilize this with. I don't know how these translators' programming is fed, but if it's retransmitting an off-air feed you'd a playback workstation at every remote site, and a way to trigger local breaks (25 Hz tones?). Likely more expense with little ROI. More work for the traffic person too, with more logs to reconcile not to mention having remote sales staff.