Isolate perf runs by protocol

[bill-ph]

Summary

• run each protocol's warmup and measured iterations as one contiguous block
• preserve declared target order and restart measured iteration numbering per protocol
• version the multi-protocol benchmark series to v3 so historical charts do not mix methodologies
• document the protocol-isolated execution contract

Why

Interleaving protocols allows a slow query on one target to leave another target idle long enough to lose its warm connection and cache context. That makes later measurements depend on unrelated protocol latency rather than the system under test.

Testing

• just test-perf
• just lint

[github-actions]

Test Impact Plan

Deterministic summary of how this PR changes tests, CI runners, and coverage-risk signals.

Summary

Area	Added	Changed	Deleted
Test files	0	5	0
E2E/journey files	0	0	0
Workflow files	0	0	0

Signals

• Test cases: +1 / -0
• Assertions: +5 / -1
• Skips or known failures added: 0
• Workflow continue-on-error added: 0
• Workflow path filters added: 0
• Test commands removed from justfile: 0
• E2E/journey retry lines added: 0

Coverage risk: neutral or increased

No coverage-reduction warnings detected.