Friend of mine has a lab, had an employee run positive grid plate samples from 3 different AGM batteries, CSB, BB, Genesis. we have more to run.
Won't have the calcium standard for the ICP machine till the end of this week but so far there is zero/zilch antimony in the remains of the positive grid plate, when it is dissolved in nitric acid and hydrogen peroxide.
did get results for copper in one battery but not so much in the others, but until a better method is developed the results are suspect.
anyhow it seems consistently some cheap agm's last 3 years others last as long as 7, when floated continuously at 13.8 volts. but the results don't really have a pattern.
one method i want to try is to setup a relay and a diode, such that the relay opens and the battery is at rest, but can still discharge into the UPS. when the power goes out, so will the timer opening the relay, so it will charge the battery, for say 2 days. then float for say, 28 days.