Counting those Test Cases

So we are working through a test strategy for a big software rollout at work, change of ERP for those of you that are interested.
As you will have read in my earlier posts we have a consultant in to help us with this, He's done a great job of highlighting some pitfalls we have overlooked, and suggested some thing that we really need to be considering. However he also has a huge amount of work in his overview for writing test cases, writing test scripts , then running those test scripts. I have challenged him on this, I have suggested that we would be much better of if we used Exploratory Testing, and get the SME's (subject matter expert) to use this rather than us writing thousands of test cases which, mostly will need rewriting as soon as the business gets a good look at the software and changes a load of stuff, like they normally do.
He seemed bemused that I'd want to take this approach on such an important project. I did my best to explain the benefits of the context driven approach, but I'm new to the whole thing, so we will see how well I do when we attempt to restructure the plan today.
Any pointers on how to sway people are gratefully accepted.