Decouple stats daemon and preserve final mine OSD status

- Run `subminer stats -b` as a dedicated daemon process, independent from the overlay app
- Stop Anki progress spinner before showing final `✓`/`x` mine result so it is not overwritten
- Keep grammar/noise subtitle tokens hoverable while stripping annotation metadata

docs-site/subtitle-annotations.md

@@ -4,6 +4,8 @@ SubMiner annotates subtitle tokens in real time as they appear in the overlay. F
 
 All four are opt-in and configured under `subtitleStyle`, `ankiConnect.knownWords`, and `ankiConnect.nPlusOne` in your config. They apply independently — you can enable any combination.
 
+Before any of those layers render, SubMiner strips annotation metadata from tokens that are usually just subtitle glue or annotation noise. Standalone particles, auxiliaries, adnominals, common explanatory endings like `んです` / `のだ`, merged trailing quote-particle forms like `...って`, repeated kana interjections, and similar non-lexical helper tokens remain hoverable in the subtitle text, but they render as plain tokens without known-word, N+1, frequency, JLPT, or name-match annotation styling.
+
 ## N+1 Word Highlighting
 
 N+1 highlighting identifies sentences where you know every word except one, making them ideal mining targets. When enabled, SubMiner builds a local cache of your known vocabulary from Anki and highlights tokens accordingly.
@@ -80,6 +82,10 @@ When `sourcePath` is omitted, SubMiner searches default install/runtime location
 Frequency highlighting skips tokens that look like non-lexical noise (kana reduplication, short kana endings like `っ`), even when dictionary ranks exist.
 :::
 
+::: info
+Frequency, JLPT, and N+1 metadata are only shown for tokens that survive the subtitle-annotation noise filter. Standalone grammar tokens like `は`, `です`, and `この` are intentionally left unannotated even if a dictionary can assign them metadata.
+:::
+
 ## JLPT Tagging
 
 JLPT tagging adds colored underlines to tokens based on their JLPT level (N1–N5), giving you an at-a-glance sense of difficulty distribution in each subtitle line.