And for each of those many ideas that are currently infeasible, there are probably other solutions that will turn out to be more practical. Technology doesn't progress by someone saying "Here's a way we can probably do this thing in the future, we just need to fill in the details." It works by fulfilling incremental goals toward improving existing technology. Any speculation about currently-impossible means to the end of overarching goals is just adding burdensome details. ( http://lesswrong.com/lw/jk/burdensome_details/ )

So if someone predicts that in the next few decades, we may have small drones that can handle gusts of wind, that's not a bad prediction. When they predict that wind gusts will be handled by a nature-inspired wing-flapping-with-feedback system the which does not remotely exist, you should stop listening. Those extra details sound like they give the hypothesis more credibility, but in fact they are constraining its predictions to a laughably tiny space.