Tracking Issue: random access Lance parity

[myrrc]

Lance is better on feature vectors by around 3-15x.

1. No selection usage in repeated reads, solved by Use selection in repeated scans #8137 (2x improvement)
2. Most requests produce small splits where BTree overhead dominates Use Vec<u64> instead of BTreeSet for splits #8194 (1.3x improvement)
3. Re-parsing flatbuffers and reinitializing chunk offsets is an overhead ViewedLayoutChildren child layout cache #8234 (likely 2x improvement
4. Chunked layout children are not cached Chunk reader children cache #8209 Allow caching LayoutReader in VortexFile #8244
5. Intermediate materialized structs deallocation.
6. io_uring et al. - we're not comparing same IO because our IO is blocking. The same issue is lock contention because every pread() is spawnedas a blocking task. Possible solution for this is vectorised reads/multi-reads.
7. feature-vectors/correlated is faster with footer reopen. Why?

Marginal improvements:

• For small split read tasks the time of reading is marginal compared to tokio task planning and LazyScanStream initialization. This may be solved by a heuristic - only for the main thread.

[myrrc]

(2) is not due to ArrayStats Arc allocation: changing Arc -> OnceLock didn't have any impact.