Thanks for reporting this.
Yes, you are right that this definition should probably be removed. I think I added it initially to force having a definition for Dagger on every tensorial symbol. But it may become problematic, as you say.
Information changed in WL a few versions ago, to return an InformationData object containing useful computable information (in an association), instead of just printing definitions.
The problem you are seeing actually comes from an evaluation leak in the typesetting of LieD derivatives by xTensor, affecting this downvalue:
Dagger[LieD[v_][expr_]] :> LieD[Dagger[v]][Dagger[expr]]
I'll fix this for the next version of xTensor.
Cheers,
Jose.