fix: exclude __words__ keyword index from corruption detection and getStats()

The __words__ keyword index stores 50-5000 entries per entity (one per
word), which inflated avg entries/entity well above the corruption
threshold of 100. This caused:

1. validateConsistency() to falsely detect corruption on every startup,
   triggering unnecessary clearAllIndexData() + rebuild() cycles
2. getStats() to log false "Metadata index may be corrupted" warnings
   and report inflated totalEntries/totalIds stats

Both methods now skip __words__ when counting, so stats and health
checks reflect metadata fields only (noun, type, createdAt, etc.).
Keyword search is unaffected since the __words__ field index itself
is not modified.

docs/vfs/VFS_INITIALIZATION.md

@@ -1,4 +1,4 @@
-# VFS Initialization Guide (v5.1.0+)
+# VFS Initialization Guide
 
 ## Quick Start
 
@@ -30,7 +30,7 @@ await vfs.init()             // ❌ Separate initialization
 await vfs.writeFile(...)
 ```
 
-### After (v5.1.0+):
+### After:
 ```javascript
 const brain = new Brainy(...)
 await brain.init()           // VFS auto-initialized!
@@ -53,7 +53,7 @@ await vfs.init()
 await vfs.writeFile('/test.txt', 'data')
 ```
 
-### New Pattern (v5.1.0+):
+### New Pattern:
 ```javascript
 // Just remove the () and init() call
 await brain.vfs.writeFile('/test.txt', 'data')
@@ -142,7 +142,7 @@ await brain.init()  // Required!
 await brain.vfs.writeFile(...)  // Now this works
 ```
 
-## Fork Support (v5.0.0+)
+## Fork Support
 
 VFS works seamlessly with Brainy's Copy-on-Write (COW) fork feature: