What can the Myo armband actually do?

This was a very thoughtful and interesting post.

It's funny, because despite the potential, it seems people these days don't care about the "what will it do for me tomorrow" as much as "what can it do for me right now."

Companies like Apple have the luxury of war chests that allow them to take on substantial risk for iterating on a hardware product in an unproven and new direction where the true potential is not able to be realized yet because of how ground breaking the device is.

I have high hopes for the Myo and anything else that can translate my movements into signals for various things. Ever since I read Rainbow's End and how people's slight twitches and movements were used as input devices for their wearable computers, I realized this is where things are going.

Someone needs to take the risk to bring this technology to market, and with Myo's vision, my hope is that they survive long enough to see their vision become reality vs. just serve as a stepping stone for another larger tech giant to make it mainstream.

Unfortunately thalmic labs is very developer unfriendly, so while the Myo still might have potential, working with the closed source code makes it very difficult to expand the functionality, especially in linux. I suspect it's just another expensive paper weight.

The short answer - it's not accurate or reliable enough to do anything useful. It is just an expensive toy and worse, artificially difficult to hack. (the raw data is restricted, even to developers)

I am curious to know the current limitations to classifying other hand gestures, such as a thumb-to-middle-finger gesture. Is the limitation due to processing of (possibly alias) signals?