Revert NSA MQA logits chunking while performance is unresolved

Temporarily revert the syh MQA-logits row-chunking port because the observed
runtime performance is worse on the current bs>1 CP prefill workload. Keep the
history explicit so the optimization can be revisited after profiling identifies
where the extra overhead comes from.

This reverts commit 4e49751406 (Bound NSA MQA logits peak memory).

Constraint: Current priority is restoring the faster known path for remote ETE runs
Rejected: Keep chunking behind the force/debug env only | the production auto-chunk path still changes runtime heuristics and should not stay until profiled
Confidence: high
Scope-risk: narrow
Directive: Reintroduce logits chunking only with ETE performance evidence and forced-chunk equivalence coverage
Tested: Local py_compile for environ.py and nsa_indexer.py
Not-tested: Remote ETE performance after revert
Co-authored-by: OmX <omx@oh-my-codex.dev>

python/sglang/srt/environ.py

@@ -220,11 +220,6 @@ class Envs:
     # large bs) but coarser overlap. 1 = per-layer.
     SGLANG_CP_SHARED_KV_PER_LAYER_GROUP = EnvInt(8)
     SGLANG_CP_SHARED_KV_USE_TAI_MATERIALIZE = EnvBool(False)
-    # NSA paged MQA-logits chunking equivalence test: when >0, force the paged
-    # topk path to chunk at this many query rows AND assert the chunked topk_result
-    # is byte-identical to the unchunked single-call result. For validation only
-    # (run a small batch so the unchunked reference fits); 0 = off (production).
-    SGLANG_NSA_MQA_LOGITS_CHUNK_FORCE_ROWS = EnvInt(0)
     SGLANG_CP_SHARED_KV_FUSED_MLA_STORE = EnvBool(False)
     SGLANG_CP_SHARED_KV_FUSED_INDEX_MQA_PREPARE = EnvBool(False)
     SGLANG_CP_SHARED_KV_ENABLE_MLA_PREFETCH = EnvBool(False)