How does Elm Virtual DOM reconciliation handle massive nested lists with keys?

The Elm runtime utilizes a Virtual DOM diffing algorithm to minimize expensive browser updates. While this architecture ensures high performance for standard state transitions, the efficiency of the reconciliation process can vary when dealing with complex, deeply nested data structures.

When using Html.mapList, the keys function is intended to help the runtime identify elements across renders. However, there is uncertainty regarding the specific performance overhead introduced when these keys are applied to massive nested lists that undergo frequent reordering or high-frequency state updates.

What is the measurable performance impact of Html.keys on reconciliation cycles in deeply nested trees? Does manual key management effectively prevent unnecessary re-renders in scenarios where the runtime cannot optimize automatically?