ERTS24 has a submission model where authors first submit a 4-page abstract. If accepted, a full version of the paper is prepared with supervision from a dedicated TPC member.
I see a number of advantages, in particular with involved efforts:
- if the paper is rejected, no full version had to be prepared
- if accepted, a TPC member helps to put the paper in the shape and form required by the conference. So, instead of the trial-and-error process of authors preparing full versions upon rejection, interpreting reviews, and guessing how the paper should look like, the conference explicitly help
- Good contributions in incorrect paper shape are supported to get published.
- Technical contributions from “outside”, not familiar with the conference, are easier to be appreciated and published at the conference.
Authors and conference, and community, win.
The total effort for the PC goes up with shepherding, but I guess it might be balanced by reduced reviewing efforts for the extended abstracts only. For that to work, shepherds have to take the job seriously, which is an effort.