FWIW, I've found the same thing on all three in my FuzzMeasure measurements of my XLS12s. I believe HobbyHiFi or K+T's results were similar on one of the drivers, though I can't find the magazine.thread from the dead ... sorry if this has been covered?
But why does the Klippel data show a voicecoil inductance that's one third the value reported on the datasheet?
It makes sense to me that the spec sheet is wrong, because frankly I can't imagine a woofer with the spec-sheet Le of the XLS series would sound as good as these do. But why Tymphany hasn't updated their spec sheets to reflect a reality that is very favorable to them, I have no idea.