As the climate modeling community approaches the time when a new experimental design for future climate simulations will be proposed to guide the next generation of coordinated coupled model experiments (CMIP Phase 6) we explore differences among scenarios based on the available experiments from CMIP5 (mostly through results from idealized 1% CO2 scenarios, but also checking for consistency of the results through historical and RCP4.5 simulations). Our two goals are achieved with the same methodological approach, despite being motivated by opposite concerns: on the one hand, we ask what level of radiative forcing—or global average temperature change—causes climate to change significantly enough to warrant exploring the underlying scenario through coupled model simulations. Significantly enough—or different enough—is here defined in terms of significantly different (both statistically and in magnitude) values of local temperature at a large-enough fraction of grid-points over the Earth's surface. On the other hand the opposite concern may be answered by our results: what level of radiative forcing is small enough to prevent climate outcomes from being significantly (by the same definition as above) different from a baseline? The first type of answers can inform the choice of a set of scenarios to be run by CMIP6 participating ESMs, avoiding costly redundancies given the computational resources required to run these simulations. The second type can allow policy relevant explorations of alternative mitigation choices, in particular assessing the consequences of stabilization paths at slightly different target levels and of differently sized overshoots. It can also provide guidance when conducting impact analysis by matching pre-existing RCP-driven outcomes to new socio-economic scenarios. The latter may imply slightly different radiative forcing but our analysis could be used to decide if still comparable climate outcomes could be assumed.