Cloud migration is one of the few enterprise programmes with an unambiguous completion criterion — the workloads are either running there or they are not — and that clarity is part of the problem. It makes migration the goal, and migration is the half that does not deliver the benefit.
The benefits in the business case were elasticity, delivery pace and cost efficiency. All three come from how the estate is operated afterwards: whether teams can provision safely without a ticket, whether resources have owners, whether spend is visible to the people who cause it. A lift-and-shift with the previous operating model attached delivers none of them, at a higher run rate than the data centre.