There isn't much matter to interact with, but not quite zero either. Something like 1e-23 of the distance had some amount of matter in it (mostly very diffuse gas clouds, nothing more).
Edit: And I didn't see Animats' comment that EM radiation was delayed ~2 seconds. But also that the speed difference was very small.
It's entirely possible that displacement in these dimensions corresponds to an increase in energy in the system that offsets the "lost" energy.
We can't say "we're definitely not experiencing displacement in X dimensions" if we don't definitively know the rules of displacement are homogeneous in every dimension. Since we don't definitively know the number of dimensions, we also can't exhaustively test these rules.
We can be more or less confident about how stuff works though. For instance, thermodynamics is something we are very confident of having gotten right, because there are plenty of ways to test if doesn't work, and those test fail all the time (so far, at least!).
So this article basically says: would it be easier to explain what we see if there were other dimensions? and the answer is it wouldn't, because we'd have to assume that things that we are very confident about (thermodynamics) are not correct, while simultaneously not providing any way to gain confidence in this new hypothesis (that there are new dimensions with different thermodynamic laws).
This might change some day: before spectrography was a thing, all the theories about star composition were non-falsiable either!!
Do they really mean to say "followed by detections"? Don't gravitational waves propagate at c? Also, I'm puzzled by "weaker than expected". One issue is calculating expected intensities for gravitational waves vs electromagnetic radiation. The other is calibration of LIGO signals. It's not obvious that resulting uncertainties are small enough for drawing reliable conclusions about gravity leakage.