Some feedback based on my experience. And some questions to the rest.
Fibaro's ball sensor will not hold in its support unless the support is mounted on real walls. On plastic, platter frames that can vibrate with the shock it will fall. I faced such a situation, while one of mines is mounted on a door frame, fixed in a framed platter wall, and it was just the wind that forced the door to close and maker the frame to vibrate. I assume the same on plastic windows frames. Meanwhile, more surprising, on Fibaro, side, I also noticed unpredictable/not desirable shock notifications (especially not on the spaces I wanted to monitor by vibration, hence I arrived disabling vibration detection on all of them: about 15 or 20).
Hence, for shock detections on my outside facing spaces (windows, doors that mattered), I added, Vision's sensors. After a month now, installed on big windows, door frames like these:
http://www.internorm.com/fr/produits.html, and all kind of shock tests I have done, I can say they react predictably only when required.
Meanwhile, as the rest of the forum stated, Vision's sensors seems to report well the shocks, but I can get them to behave decently to report wake up; or is it our BOX? I appreciate feedback if anybody made progress in this respect. Like I said, I am not even sure if the problem is in the sensors, while excluding/including over multiple concrete floors, reporting of shocks is all good. Even reporting of programmed wake up time is right. But no ACK in the BOX.
Maison avec eedomus+, 200 dispositifs(60 en zwave/+, 10 en RFXcom, 5 cameras, les restes virtuel), automatisation éclairage, sécurité maison, environnement, access, chauffage, notifications, agenda.