Why, after more than 20 years, is the notion that IR35 status determination is a mystical art, with an infinite list of variables and combinations that warrant a super computer in order to fathom, still in play? It's a tiny data set. Unless you unnecessarily expand it to make it look big. Surely after 20 years, we can figure out a combination of working practices for which we know the result? Assessment models know the criteria. They just hide them in a black box. And working practices are not a fixed set that clients, agents and contractors can't derive, or evolve, or refine, or design. Or just pick. Simply implement the working practices and contractual relationship that represent the best all round benefits case.
There will be engagements that may be inescapably borderline, perhaps if there is no appetite for a right of substitution, or feasibility to execute a substitution. Even in those cases, it can be straightforward, based upon probably less than 10 key factors, to derive working practices that should indisputably drive a predictable IR35 status.
How can it be that we have lots of different tools out there, and we'd get different results for the same essential inputs? That’s the REAL mystery.