Still not persuaded I'm afraid. You want the helmet to compress in the bit between the point of contact of your head and the surface it hits. You surely really don't care what happens elsewhere.
There will absolutely be scenarios where a helmet cracks in that place or close to it and thereby impairs its ability to compress. But there are also scenarios where the crack is elsewhere. I suspect one scenario is that a helmet is a loose fit on the head. The first effect of an impact is to bend the bit of the helmet that hits first closer to the head, a bend that creates cracks at the point of the bending (or, if I am allowed a medium length word, at the nearest stress raiser). If those cracks are to the sides of the impact point, I don't see that it has any effect on the ability of the bit of polystyrene in that crucial area to absorb energy by compressing.
I am no blind advocate of helmets- I have ruined at least one dinner party by not keeping quiet and I've stopped wearing one routinely myself - but I think there are times when this thread succumbs to the "leaning into the wind" syndrome - in our zeal to counter extravagant claims in favour, we overstate the arguments against.
There are a number of significant factors that need to be considered. When I say "lateral movement", this is not merely seperation between the two crack surfaces. This also includes movement where both surfaces move past each other... err... laterally (both normal and orthogonally to the impact force vector). Consider that the helmet is a (roughly) spherical structure:
it is inevitable that an impact will result in substantial shear forces being set up some distance away from the impact site. Shear forces are the principal means by which the two crack surfaces move past each other. Furthermore, given that the foam is an elastic material, once the load is removed the crack will spring back, at least to a degree.
It is therefore impossible to judge from the seperation of the crack afterwards how much movement took place during impact. This is important - compression takes up at most 20 mm. Even if movement along a crack is only 2 mm, it can be readily seen that it is sufficient to materially diminish the protection offered.
It is not true that cracks distal from the impact site are unimportant. The essential thing about a crack is that
it will propagate most readily in regions of maximum stress. A crack that starts some way from the impact is unlikely to stay there - it is highly energetically favourable for it to propagate
to the impact site. Remember that a helmet is an integral load bearing structure and even elements far away have a direct bearing on its strength. I referred to this above, by pointing out that shear stress is distributed throughout the structure. And this happens whether or not the helmet fits properly.
Which brings us to stress raisers. (I deleted a post about this the other night, thinking that it was a little too arcane. Ha!) It has taken a long time for it to be understood just how important shape is - think of say the Liberty ships, where some broke in half because a crack started from the sharp corner of a hatch. Or the Comet crashes: the designers were no fools and reinforced the aluminium round the windows, but didn't understand the sharp corners amplified the forces in that region by an order of magnitude. In short, the sharper the corner, the higher the stress amplification. It is disturbing how often sharp corners are seen in helmet vents. Impact stresses in those areas can be expected to far higher than the rest of the helmet. It is not a sign of good design - I wonder just how much the designers really understand about the way materials fail.