Sorta works but deadlocks on larger inputs.

Still doesn't fix deadlocks tho :/

Still WIP

Fastest results on AMD at workgroup = 1024. Note, this has atomicOr
workaround for correctness.

Also note, not all targets will support a workgroup of this size; on
shipping, we'd need to query and select at runtime.

Do a small sequential scan at the leaf of the hierarchy. That amortizes
both the workgroup-scope tree reduction and the (still sequential)
decoupled look-back to a larger number of inputs.

Note: this falls short of a real performance evaluation because there's
no attempt to warm up the GPU clock. But it's valid as a very rough
swag.

Better for performance analaysis

Performance measurement requires keeping the GPU busy. That means not
copying results back to CPU and doing verification there.

Naga will accept ordinary loads and stores to atomic types, but tint
will not.