That is one of those Wikipedia articles which is a bit vague about what it means. It's doesn't make sense to intend to say that glass transmits 90% of incident light regardless of the thickness. The Wikipedia entry references a single optical "element", so I'd take "the transmissivity of one element (two surfaces) is about 90%," to mean that 10% is the lower limit of light loss for a single lens of arbitrary thinness.
Now if a very thin silica glass lens transmits 90% of the light falling on it, then clearly it'd be very difficult to conceive of a material that transmits 10% more light than that. However you can achieve whatever level of attenuation you wish by making your piece of glass sufficiently (possibly absurdly) thick. The three inch thick glass panes used in giant ocean tanks are noticeably more opaque than air. Clearly it's physically possible for a material to transmit 10% more light than the same thickness of glass -- for a sufficient thickness. Particularly if the index of refraction of that material is closer to air.
Of course that's where we get to the point that the summary is badly written too. Silica glass *is* very transparent; insufficient transparency isn't a problem in window applications, if there's a problem it's that the material is too transparent. That's why we have dark tinting and anti-IR coating. So it's not clear why we would care that the material can transmit 10% more light. Clearly the story got garbled somewhere along the way.