Gullanian is absolutely right - hell, if a thousand people put in $10 each, I'd start work on an Android native exporter right away! HTML5 can do it in a browser, but Android native does have some advantages: you can run it offline, from the phone itself, without going through a browser; it can go fullscreen; you can probably use OpenGL ES and all its features; you can access the phone's native functionality which you can't get through a browser. Same for iOS. HTML5 is a quick way to get basic platform coverage, there are still some advantages to running native (just like the desktop runtime would have advantages due to running natively). So it's a perfectly valid suggestion. This is how it's supposed to work - someone wants that, and now other people can pledge funds towards it too. It's a lot of work though, so would probably need a lot more pledges to attract developer attention than, say, a smaller feature.
I don't see what you mean about GPL... can't see a problem there. If the developer wants to sell their work after receiving all the pledge money, that's a bit trickier - however, for a smaller unit of work like a plugin, the pledge hopefully would serve as a fair one-time payment for the development work.