Dan,
Similiar here. Many times questioned.
- Alarm status off greeting in the house is personal relating to whom shut off the alarm. (historically have added house status, temperature, messages etc)
- hourly chimes which include the outside and inside temperatures (brother in law asked why?)
- certain outdoor motion zones (side parts of house - doors or windows)
- mail delivery
- doorbell (keep flipping from basic chimes to chimes and TTS)
- CNN / BBC screen scrape "hot" news alerts
- weather alerts (again on and off again with TTS)
- garage TTS as above and also reading the RFID tag of the vehicle and announcing which vehicle is approaching or leaving
- loss of my time server (using one box, GPS for internal time NTP)
Some folks I recall get a bit carried away with a TTS syopsis of past motion activities in their house using combinations of sensors. (a bit paranoid to
When first using TTS in the 1980's I had speech associated with every zone (even doors and windows) mostly because of my kids (infants at the time). It was only one voice font. It was a real voice font versus the early 80's machine like TTS voice font. Today I have a "library" of various speech fonts (8 or more I believe) but have stuck with one over the last 10 years. (Neospeech Kate). Some folks have the TTS voices talk to each. Many times when traveling I would "text" message TTS to the house to just say hi to my wife. I tried over the years using VR but lost the "training" VR patience after a bit. But I do see in the future of HA a totally interactive touch screen with TTS/VR and even HA virtual entity interaction.
Sometimes I use the TTS announcements as a guage on how much work the HA server is doing (it used to be if it was working very hard like in the days of 500Mhz / 512mb of memory PC's and using the first AT&T voice fonts - which were already memory/utilization hogs).