This wouldn't happen if the initial allocation of Glass units had been done in sane way, like setting a market-clearing price, rather than basing it on Twitter follower count.
Did you see the list of people who got them? A significant number had no followers at all. I doubt the median follower count of selected accounts was much different from the median follower accounts of the active twitter population.
> setting a market-clearing price
That would be _way_ less of a random sampling than what they have now. You don't get good feedback by having it available to such a tiny segments of the population.
It makes sense if you look at this as a promotion/market research instead of a first batch of sales. And I think that's exactly what they're doing, giving the product to a large group of people who can reach a larger number of people to introduce their product.
It's also a bit of market and use testing since they'll be able to see how people are using them and improve or expand the application offerings in those areas.
(And yes, I'm pissed that I didn't get one.)