Do you have more detail on the attitude determination algorithm? I haven't heard of that being used before, but as an engineer I find it interesting. I understand if there are concerns about proprietary methods or ITAR.
Mostly for proprietary reasons we don't go into deep details - the algorithmic fusion of the rate sensors, accelerometers, and other data is part of the secret sauce that yields an attitude that actually works (take your iphone/ipad/android flying with any of the apps that have access to the same sorts of data to see just how bad bad can be).
But, conceptually, force (accelerometers) alone isn't enough to know which way is up, because if you're perfectly coordinated, you always think down is through the seat of your pants, right? So then you need to know how fast you're pitching/rolling/yawing. That your gyros (rate sensors in EFIS-speak). But, all rate sensors drift. In military aircraft and commercial airliners you minimize drift with money: they use sensors that can cost more than our whole panel. In GA (including all certified panels and EFISs in GA that you're familiar with), you use "aiding" from different sources, along with fancy math, to tease out what motion is sensor drift and what is real movement. We use airspeed as our primary aid and GPS as a backup. Other systems use some combination of those and add magnetic heading as well. One strength of our approach is that it doesn't require that you sit still for any length of time doing an initial alignment. In fact, you can turn on or reboot our products in the air, and after a few seconds of straight/level flight, you'll have a fully-functioning attitude indicator. Others may require a warm-up period where you have to wait and keep the aircraft completely still upon initial system power-up. Some may be completely in-op or unreliable if you happened to have any power glitch or other event that caused them to reboot. To toot our own horn just a little more, if you equip our EFISes with a backup battery, even this is REALLY hard to do