The problem with this is simple. We need to re-assess what needs to be done and then it may be the case that we need to cull some in-flight stuff. If we're culling a £1m project that's 1/4 through, we're not wasting £250k but saving £750k overall. We have still spent £250k but that's better than spending the whole £1m on what's now the wrong thing. These are tough choices.
We need to consider how we're going to implement the new model. Perhaps the new model is wrong...we need to test that's right before we implement anything, otherwise cash is wasted. We then need to run a pilot or at least walkthrough using different scenarios, or perhaps run some more complex analysis, or we're just running into the dark.