I guess this fits well in real options theory, but I would be wary of using a technique like this.

It banks on security by obscurity (that you’re throwing in the “no brown M&Ms” rule as a light-weight check), because if the team knows where those checks are, it would render them useless.

A more useful real option to reconsider the project would be to track a metric — or, preferably, set of metrics — and use thresholds on it to determine when that discussion is appropriate. This way, there’s no need to conceal any information from the team, and those metrics can be used for other stuff as well (reporting to investors, for instance).