The issue is not, I think, one of narrow-mindedness. Most linguists are interested in theoretical frameworks of understanding how languages work, and what they do in the wild. Conlangs don't really help with this. Although Lojban and Toki Pona and a few others have interesting linguistic facts, they nonetheless fail to provide insight into what language is and how it works and develops* (since, by definition, they have been designed and not developed naturally). And, since most conlangs completely lack anything of linguistic interest, it isn't surprising that the few that do tend to be missed.What's more, most conlangs (again with a few notable exceptions) look very, very much like the native language of their designer(s). Lojban is included in this category, by the way: in terms of word order, general morphological processes, etc., Lojban looks very much like a nicely behaved Indo-European language, albeit one with extensions for the aspects of predicate logic. There isn't much reason to look at conlangs for linguistic evidence or theory (although sociolinguistic aspects and other things could certainly be of interest).