The same problem was with early car airbags. The issue is long resolved.
(Perhaps you can make airbag-helmets that inflate when the bike hits something, which will do for the case where you hit something in the road and get thrown off, or where a car collides with the bike. Not so good for tree branches and such, though.)
By any measure, it sure beats riding without helmet.
Obviously an airbag helmet will be safer (but more inconvenient) than no helmet. Presumably it will be less safe (but maybe more convenient and better looking) than an ordinary helmet.
It needs to beat both to succeed: that is, either some people who currently use helmets need to be convinced that the loss in safety from using an airbag helmet is small enough to be outweighed by the gain in convenience or looks or whatever, or some people who currently don't use helmets need to be convinced that the gain in safety from using an airbag helmet is large enough not to be outweighed by the loss in convenience or looks or whatever.
This seems pretty difficult to me, especially as I bet the convenience gain isn't very big: I bet an effective airbag helmet needs to fit you about as well as an ordinary helmet, and it'll need to be put on with about as much care.
In that video where the rider was hit from behind, there was very low relative change in velocity of the sensors in that helmet - no direct impact spike - and its on par with the same levels of acceleration you would experience in everyday bike riding.
Also, I'm sure in the 5 years they spent on product R&D, the concerns that came up to you in few minutes here were raised too.
The accelerometers in your car are bolted to your 4,000lb frame, riding on dampened shocks and rubber tires - you will never experience acceleration anywhere close to what you would in a frame-on collision.
They don't have a shipping product yet.