It's not sending someone to the moon, but it's also not worthless. The system itself would benefit from hard technical solutions as well. Facial recognition in gifs; understanding context in a picture... you may see it as a waste, but it may also lead to discovery some new CS techniques for these things.
I'm not necessarily saying they can show a value of 600m, but I would just dismiss the whole idea either.
If at least this worked half of the time, I may say it is valuable.
Other sites (Google image search, for example) let me look at a single frame static preview, and only look at the animated versions of the gifs that look promising. That works quite a bit better for me. It's even faster, since it's hard to browse anything when your browser's become unresponsive.
They may have captured the hearts of some segments of the market, but things like that leave me thinking they've got a long way to go before wider appeal is a realistic possibility. Meanwhile, challengers are sprouting like mushrooms.
It's really interesting, and also when I began to feel old. Snapchat filters, the popularity of livestreaming, 50% or more of message stream being gifs, etc.