Driver breakin period, what’s the science?

So have these new speakers and been told they need a hundred hours to be broken in, and then sound will improve.

What’s going on as break in occurs?  More important for tweet, mid or bass?  
My initial listening has simple vocals/music passages sounding very good, and more complex and very layered sections that may have potential to improve.  
No one has even attempted to answer the question: What is the science?

The two parts that are easiest to explain are spiders and surrounds. These are typically accordion shaped or pleated, folded, however you want to call it, and so they stretch and bend with speaker cone travel. Whatever the material, almost doesn’t even matter, it is never perfectly manufactured. There are always lots of little regions of stress and tension or compression. Playing music stretches and compresses the material. In some of these restricting areas the bonds break and this reduces stiction and the driver moves more freely.

Think of a crease in a sheet of paper. Fold it back and forth a few times, the crease that was stiff becomes almost like a hinge. Might not be measurable but that freedom of movement allows the driver to respond to big dynamic swings and subtle audio details. Exactly what we hear with break-in.

Break-in of speakers is not a "pie in the sky," but the phenomenon and the time it’ll take for the speaker to settle can vary of lot depending on playback level and the specific drivers and passive cross-over parts used. Alan Shaw of Harbeth will tell you it’s bogus and a phycological effect, and if that’s a universal statement meant to encompass most every speaker, I’d disagree. I don’t know whether Harbeth tests all of their speakers for a duration of time before they leave the factory, but if so I wager the costumer buying their speakers will notice less of a difference with regard to break-in. This may of course apply to other brands as well. Even so, if speakers have been stored up for a while, not least in low-ish temperatures, they’ll take a little time to settle in - despite the fact that they may have been tested for hours prior to leaving the factory/assembly facility. More mechanically rigid, rugged suspension materials like pleated cloth surrounds, typically of higher sensitivity drivers, as well as their spiders are prone to need a beating before loosening up (see the 6moons review of the WLM Diva Monitor with a 10" Eminence coaxial driver).

(more speculative, perhaps)
While the mechanical, moving parts of drivers may account for the biggest change in sound during break-in, there’s also the possible influence of magnetization of the areas surrounding the magnet. For this unwanted magnetization to cease (permanently) requires heat build-up in this voice coil and magnet, in other words that the speakers have been played at fairly loud (though still safe) volumes for a duration of time. Maybe this accounts for why some people feel their speakers suddenly sound better, even after years of use, after a party or other thorough workout they haven’t previously received.

Even though my current speakers have been broken-in (from years of actual pro cinema use), I find they sound better when played at moderate to loud levels for an hour or more (or less, depending again of the playback level) after I turn on my stereo. A well known scenario is what follows having watched a loud action movie for a couple of hours, where they sound their very best.
It really does depend upon the type of speaker. Planers especially benefit from a break-in period. But mostly it's psycho-acoustics in that it takes time for the listener to convince himself that his speaker purchase was money well spent.