fix: dynamic Windows VM boot synchronization for RDP

[felixzsh]

The current 5-second sleep is not enough for all hardware. On my PC (and likely many others), Windows takes longer to initialize. If the script tries to connect via RDP before Windows is fully ready, the connection fails or hangs.

I replaced the hard-coded sleep with a dynamic loop that checks the Docker logs for the "Windows started successfully" message. This ensures the RDP client only starts once Windows has confirmed it's ready, making the launch process much more reliable across different hardware specs.

Fixes #2599

[felixzsh]

for further revision, i can see how the RDP port and the windows vm loops could be unified like this:

    echo "waiting for windows vm to start..."
    WAIT_COUNT=0
    # check RDP port
    until nc -z 127.0.0.1 3389 2>/dev/null && \
          # check for windows start message
          docker logs omarchy-windows 2>&1 | grep -qi "windows started successfully"; do
      sleep 2
      WAIT_COUNT=$((WAIT_COUNT + 1))
      if [ $WAIT_COUNT -gt 60 ]; then # 2 minutes timeout
        echo "❌ timeout: windows vm failed to start within 2 minutes"
        echo "   the vm might still be installing windows."
        echo "   check progress at: http://127.0.0.1:8006"
        echo "   check logs: docker logs omarchy-windows"
        exit 1
      fi
    done

---

but i dont know what is the best, clean and lees code, or separated timeout errors