There's a lot of value in periodically reassessing your assumptions. It might not always change your mind, but occasionally you'll find that things have shifted enough to alter long-standing conclusions. I'm at such a point with regard to hydrogen-powered internal combustion engines. BMW is introducing their second-generation hydrogen 7-Series--still not a mass-production model, but a bit closer, with the ability to switch between gasoline and H2. As this article in MIT's Technology Review reminds us, its environmental benefits hinge on how its hydrogen supply is made, and that's been my problem with this concept for many years. Turning natural gas into hydrogen--the primary production method today--and then burning it in an ICE is a pretty dodgy proposition, both economically and environmentally. If that were the whole story, my judgment would remain unchanged: bad idea.
What has changed since the last time I gave this proposition much thought is the prospect of large amounts of intermittent renewable electricity. Before, the best use of that electricity always seemed to be in backing down our use of those electricity sources that contribute the most local pollution and greenhouse gas emissions, typically coal-fired power plants. But if there's more wind power than the local grid can absorb, turning the surplus into hydrogen for cars like the new BMW could make sense.
Let's be clear: an ICE is never going to use hydrogen as efficiently as a fuel cell car would, and many of the advocates of H2 ICEs see them only as a bridge to encourage the necessary refueling infrastructure to develop, while fuel cell developers work towards a product that's cheap enough to compete. It's a way to break the key chicken-and-egg hurdle facing fuel cells, even if BMW doesn't see it this way. In any case, demand for hydrogen in transportation is likely to be modest for many years to come.
Now, in the real world there would be no way to be sure where the H2 fuel for cars like this will come from, day to day, just as you can't be certain that the gasoline you buy won't come from oil produced in Libya, or Venezuela, or some other country whose government you might dislike. But as long as some of the hydrogen came from renewable sources that are truly surplus to power generation, this ought to be good enough to overcome the technical concerns of people like me, who focus on problems that don't trouble the public very much in any case, as we've seen for years with corn ethanol.