The European DEMO fusion power reactor (EU-DEMO) is still in the pre-conceptual design phase. The design strategy for the EU-DEMO hinges on investigating multiple reactor designs and technologies in parallel, progressively down-selecting these in the mid-2020’s, in preparation for the conceptual design phase. The present implementation of the strategy centres around a baseline single-null design, which is configuration controlled and iterated approximately every two years. The majority of resources are dedicated to studying the baseline design; changing it is expensive, and takes months to do. Competing technologies for different sub-systems (e.g. blankets, magnets) are forced to co-exist within the same physical confines (CAD models), and conform to the same set of performance criteria. Meanwhile, the alternative reactor designs are only loosely defined, with no agreed set of parameters, no CAD models, and no formal framework for study. We argue that the EU-DEMO design strategy and the implementation approach are fundamentally incompatible, and are not likely to lead to a successful outcome. We argue that the strategy is correct, but the implementation is inadequate. We make the case for a change in approach to the design of DEMO-class reactors in Europe, and propose a solution which, while poles apart from the present approach, is more politically and financially feasible than the arguably ideal situation of a large, fully funded, centrally located team devoid of all national affiliations and political connections. We present our preliminary attempts to demonstrate the feasibility of our idea, in the form of a new fusion reactor design code: BLUEPRINT. We demonstrate that the typical activities required to generate a DEMO design point can be sped up by four orders of magnitude — from months to minutes— paving the way for a rigorous and broad exploration of the design space.