At that point Twitter will simply block IP address, or some such. Sure there will be a game of catch me if you can, but the end is that no one really wins, and Twitter may suffer because people will use it to bad mouth Twitter. Enough user sentiment, even if it start from the devs of a few highly used apps, and Twitter might want to reconsider.
If the client is native, you can do that scrapping client side. Say your app has 10000 users. Assuming they can detect this (not likely) what are they gonna do? Block 10000 IPs?
Translation: so that we can charge even light API users.
This will surely backfire - some services will switch from API usage to screen-scraping, resulting in an even higher load on twitter's servers.