docs(readme): drop two comparison claims the build does not deliver

"Sandboxed Code Verification" and "Spoken Answer Humanizer" were marked
✅ in both the competitor comparison and the Free-vs-Pro table. Neither
runs for a user on 2.8.7:

- codeVerification/verificationEnabled.ts documents itself as
  "TEMPORARILY DISABLED across the app", defaults OFF, and the
  orchestrator and ipcHandlers early-return when it is off.
- The humanizer is gated on `answerDiversityGuard`, which is
  `default: false` in intelligenceFlags.ts.

Neither `codeVerificationEnabled` nor `answerDiversityGuardEnabled`
appears anywhere under src/, so there is no in-app opt-in either — the
only way to turn them on is an env var and a relaunch from a shell. That
rules out downgrading to ⚠️ "partial": there is no path a user can take.
Removed rather than footnoted; they can come back when the kill switch
flips.

Left alone: Hindsight LTM (real toggles in Settings → Intelligence, so
the Pro ✅ is accurate) and Stateful Intelligence OS (the V3 answer layer
is default ON). The "Humanize AI Text" Skill is a separate, working
feature and is untouched.

Both tables re-checked for column consistency after the removals.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01JJS2xjF8uDPMa4EpXey5FL

Author: Hardening v2.7

README.md

@@ -187,8 +187,6 @@ This demo shows **a complete live meeting scenario**:
 | **Codex CLI Integration** | ✅ Yes                     | ❌                   | ❌         | ❌               | ❌                     |
 | **Offline SLM Mode**      | ✅ Yes                     | ❌                   | ❌         | ❌               | ❌                     |
 | **Stateful Intelligence OS** | ✅ Yes                    | ❌                   | ❌         | ❌               | ❌                     |
-| **Spoken Answer Humanizer** | ✅ Yes                     | ❌                   | ❌         | ❌               | ❌                     |
-| **Sandboxed Code Verification** | ✅ Yes                  | ❌                   | ❌         | ❌               | ❌                     |
 | **Hindsight LTM Vector DB** | ✅ Yes                    | ❌                   | ❌         | ❌               | ❌                     |
 | **Regional STT Relay**    | ✅ Yes                     | ❌                   | ❌         | ❌               | ❌                     |
 | **Data breach history**   | ✅ None                    | ❌ 83k users exposed | ✅ None    | ✅ None          | ✅ None                |
@@ -392,8 +390,6 @@ We've launched the official **$NAT token** on Printr! Holders who maintain a spe
 | **Undetectable & Stealth Modes**                    |      ✅       |      ✅      |
 | **Meeting Dashboard & Offline RAG History**         |      ✅       |      ✅      |
 | **Stateful "Intelligence OS"**                     |      ✅       |      ✅      |
-| **Spoken Answer Humanizer**                         |      ✅       |      ✅      |
-| **Sandboxed Code Verification**                     |      ✅       |      ✅      |
 | **Hindsight Long-Term Memory (LTM)**                |      ❌       |      ✅      |
 | **Job Description (JD) & Resume Context Awareness** |      ❌       |      ✅      |
 | **Automated Company Research & Dossiers**           |      ❌       |      ✅      |