I give a brief introduction to and explain the geometry of teleparallel models of modified gravity. In particular I explain why, in my opinion, the covariantised approaches are not needed and the Weitzenb{\"o}ck connection is the most natural representation of the parallel transport structure. An interesting point is that it also applies to the symmetric teleparallel case. I also share my thoughts on why the teleparallel framework does not seem to be a next rung in the ladder of understanding the real world's gravity. At the same time, these theories do have a clear and justified academic interest to them.