That's right: even if the dipole field diminishes, the secular field is left -- there's never a time when the field's zero. True, the non-dipole field is highly changeable, and would offer a patchy coverage, but it would still help us to ward off solar storms. And there's no evidence at all, looking at the past record of species extinctions, to statistically match up a field flip with the destruction of life. [Of course, destroying our fragile human civilisation is another thing entirely!]
And it's also true that we misinterpret the meaning of 'sudden': to a geologist, 10,000 years is almost instantaneous!
Well, the dinosaur event may not be as 'cataclysmic' as some think... There's plenty of evidence that it wasn't a sudden wipeout triggered by an asteroid, though the Chicxulub strike (and others) certainly won't have helped. Huge magma outpourings (over hundreds of thousands of years) in the Deccan Traps are also hypothesised to be linked to the extinctions.
We tend to think of 'sudden' events as taking place overnight, or in a short period by human timescales. In geological terms, 10,000 years is an eyeblink, remember!