Okay; 15x the hash power for these other devices, which are a bit cheaper per GH/s.  For these devices to lose money mining (versus electricity), we need the overall network hash rate to increase beyond 100 fold what it currently is now.  At that point, everyone is likely losing money unless someone comes out with a way more energy efficient process.  There shouldn't be any point in the next couple years where I'm losing much money mining with these.  Maybe I'll make $250 fiat bucks back, or $100, or $1000.  Doesn't really matter to me; I'm supporting the BTC network by a decentralized means.

No one can tell what the price of BTC will be in the future, and also what the network hash rate will be in the future.  The price may fall to the point where it's not profitable for large ASIC mining companies to continue their operations mining or manufacturing products (similar to what happened after the 2011 crash), the difficulty could collapse, everyone could start throwing their devices out windows because they are worthless.  I can continue mining and losing a few bucks per month with my little USB sticks at pretty much no risk to me (a cheeseburger a month in power -- who cares?).  Then we may see the price go up in the long term.  One of the best times to be mining BTC before was when it was seldom very profitable by GPU in 2012, because the difficulty kept falling and everyone still mining kept getting more and more BTC; over time the BTC proved very valuable.

This is my high risk investment.  I know I'll make at least a little back, I have no worries about the loss of $500, and I'll be securing the network in the meantime.