perf:improve performance of morton order iter

[d-v-b]

Our morton_order_iter function uses nested O(N) loops and doesn't cache its results. These make it slow. This PR fixes both of these, so it should be faster.

[codspeed-hq]

Merging this PR will improve performance by ×2.1

⚡ 4 improved benchmarks
✅ 44 untouched benchmarks
⏩ 6 skipped benchmarks¹

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	WallTime	test_write_array[memory-Layout(shape=(1000000,), chunks=(100,), shards=(1000000,))-gzip]	15.5 s	9.6 s	+61.75%
⚡	WallTime	test_write_array[memory-Layout(shape=(1000000,), chunks=(100,), shards=(1000000,))-None]	11.4 s	5.4 s	×2.1
⚡	WallTime	test_write_array[local-Layout(shape=(1000000,), chunks=(100,), shards=(1000000,))-None]	11.4 s	5.5 s	×2.1
⚡	WallTime	test_write_array[local-Layout(shape=(1000000,), chunks=(100,), shards=(1000000,))-gzip]	16 s	9.6 s	+66.84%

---

_{Comparing d-v-b:perf/improve-morton-ordering (05e16ca) with main (1ed266d)}

Footnotes

1. 6 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[normanrz]

Really cool!