1) Stop buying first drafts. While we've all heard the stories of writers who got screwed because KS had to re-write their work, one horrible implication that hasn't really sunk in is that KS pays writers for their first drafts. Any other publisher would blue-pencil that draft and send it back to the writer for a re-write rather than spend valuable executive time on that sort of work.
KS really needs to learn to be an editor, rather than a writer and designer. One of the things keeping Palladium back is the fact that so much of their output is from his sweaty fingers. If he would just let his people innovate, Palladium could keep being an industry innovator.
2) Playtest. Every other game company does internal playtesting. If a Palladium book gets playtested, it is because the writers took it upon themselves to do that. But then it gets the KS re-write anyway, invalidating about half that work.
And a writer's playtest is,. by necessity, limited. One group, maybe two if they're industrious. Enough to help the book conform to one playstyle, one set of house rules. But not enough to bang it into greatness.
Open playtesting (like Paizo does) could help Palladium go viral, but even closed playtesting with select volunteers has the potential to generate more buzz than KS can on his own.