Adding wildcards to make divisions meaningful?

It turns out I have even more company in my wildcard-makes-division-races meaningful crusade: Jayson Stark (and apparently Elias and SI's Tom Verducci). Stark, and everyone else, now recognize that when the two best teams play in the same division (Yanks-Rays this year) and both are guaranteed to make the play-offs, the incentive to win the division all but disappears, because the single benefit of home-field advantage is minimally important (Stark points out that the team without home-field advantage wins 50 % of post-season series).

The solution, according to Stark, Verducci, et al., is not to eliminate the wild card, but to add a second wild-card in each league. Now the two wild-cards play some type of play-off (he debates whether it should be a one-game winner-take-all or best-of-three and how it should be structured [Update: Tom Verducci insists it has to be a one-game playoff, not a series]) for the right to move on and play, presumably, the division winner with the best record. Now there is a genuine incentive to win the division--avoiding having to play anywhere from one to three additional games, perhaps without off-days and perhaps without a break between the wild-card series and the Division Series. And, according to Stark, people close to Bud Selig reportedly say he likes the idea.

I am not quite convinced, because it still devalues the division in non-close races. If a second-place team falls far enough behind the first-place team in its division, it turns its attention to teams in other divisions and just has to focus on staying ahead of the non-first-place teams in those divisions. So the "race" is between # 2 in the East and # 2 in the Central--although those teams will not play one another regularly in September, since the schedule is weighted towards intra-division games late in the season, for obvious reasons. Still, anything that gives a real incentive to finish first is a vast improvement.