I didn't think the concept was about having a role change capability: I thought it was about reducing development, certification and in-service support costs for subsequent variants by re-using the fuselage/engine/cockpit module of the initial trainer. The hidden armies in DE&S and industry acquiring and supporting each aircraft type are a huge cost burden on the RAF, so reducing the number of distinct types does have a certain attraction. Once you'd acquired, operated and supported some trainers, it should be much cheaper to acquire, operate and support a (eg) light fighter variant and/or a recce drone variant than it would be to acquire totally different platforms for each role.
However, the fundamental problems of modularity remain. Maybe they could find a market for such an approach in countries which have historically used light fighters for their air defence, although personally I doubt it. For the UK, the large size of our area of air defence responsibility drives range and loiter requirements which I can't see being met by an aircraft derived from an economically-viable trainer.