Agreed. I really see no reason you can't do this at any scale. We do it today in our tiny startup. The strategy we employ is a little different - Some of us "secretly" work on product iterations so as to not distract the rest of the team until we are sure we've found something. If the new work doesn't pan out, no one even hears about it. If it's amazing, we sort out a gradual introduction of the idea to the team.
There is no reason you have to have 100% of the organization bought into a rewrite/refactor/iteration of the product. All it takes is for one or 2 strong employees to go off into the dark forest for a while and come back with a potential prize. For an organization as large as Southwest, this would be a negligible cost even factoring in their razor-thin operating margins.