Currently, there are a number of proposals which seem to struggle to get into beta but which could also be subsumed in the emerging Computer Science proposal (see here). I was asking myself: when is it appropriate to suggest a merge?
More to the point, the question is: When should we consider an unclosed, not yet in beta proposal failed? As far as I am aware there is no rule/process for this.
One thing I can think of is to investigate the graph of followers/commiters over time. If it shows (significant) characteristics of saturation (i.e. convergence), start a process that checks wether the proposal is still necessary, suggests merges with other proposals (that might have created and even launched while the failed proposal was in either phase) or close it altogether.