#1 In real world, sometimes we have to make ways for urgent items. So I think we should also have a dimension of urgency. E.g. a traffic spike on black Friday needed to be addressed by releasing some patch. Some enhancement for a compliance change wasn’t planned before.
#2 I prefer a slightly different approach to deal with high impact and high effort ideas. As first step, we should quantify the effort( e.g. total hours), then look at bandwidth available to figure out how much time it might take to develop the feature. Any feature/change that might take more than 4–6 weeks of time, should be curtailed to a smaller version to test whether our estimation of impact is correct or not. Ship a lean version of the feature, measure impact and then make further investment.