Seed wait stats config file (#215 part 1)

[erikdarlingdata]

Part 1 of #215 — adds the human-readable config file that will become the single source of truth for wait stat handling.

What's in the file

src/PlanViewer.Core/Resources/WaitStats.json — 52 wait stat entries, populated only with the first 5 attributes per Joe's instruction:

1. Name
2. isPreemptive
3. isExternal
4. isImplemented
5. isEnabled

Attributes 6-12 (showWaitCount, showAverageWaitTime, timeCalculationModel, applicableOperatorNames, description, helpfulUrls, internalComment) are intentionally null, reserved for domain-expert population.

How seeding was determined

• Implemented (46 entries): every wait stat the tool today references explicitly, plus representative members of prefix patterns the tool matches (LCK_M_, PAGELATCH_, LATCH_, HT, PREEMPTIVE_*). Inventory came from BenefitScorer.ClassifyWaitType(), PlanViewerControl.GetWaitCategory(), and WaitStatsKnowledge.cs.
• Not implemented (6 entries): common waits that today fall to the "Other" bucket — THREADPOOL, BACKUPIO, BACKUPBUFFER, HADR_SYNC_COMMIT, DBMIRROR_DBM_EVENT, SQLTRACE_BUFFER_FLUSH — included so both states are represented.
• isPreemptive: true iff name starts with PREEMPTIVE_.
• isExternal: true iff the tool's IsExternalWait() routes the wait through the CPU-based external formula (today: PREEMPTIVE_* and MEMORY_ALLOCATION*). Conservative.
• isEnabled: true everywhere — the tool today only filters at the wait_category level (Idle, User Wait), and none of these entries fall into those.

Not in this PR

• No code is wired to read this file yet — that's Part 2.
• No build plumbing (<EmbeddedResource>) added yet — file is a plain checked-in JSON so reviewers can read the diff cleanly.

Test plan

• @jobbish-sql review of seed coverage (any waits worth adding/removing?)
• @jobbish-sql review of attribute-1-5 classifications

🤖 Generated with Claude Code