Founders and long-tenured leaders are rarely reluctant to delegate in principle. Most can describe, accurately, which decisions they should stop making personally. The pattern breaks down at the point of actual consequence — the first time a delegated decision goes visibly wrong, the decision quietly returns to its original owner, and it rarely leaves again.
This is not a trust problem so much as a design problem. Delegation that is not paired with a recoverable first attempt — support, a defined blast radius, a clear point at which the founder will not intervene — sets the delegate up to either avoid the decision or make it badly under pressure to prove themselves quickly.
The organisations that successfully shift decision-making down do it deliberately: they choose which decisions to hand over first based on how recoverable a mistake would be, not on how ready the person seems, and they hold the line on not re-absorbing the decision the first time it is tested.