the reason is simply because you are thinking that GeneView is somehow a prediction track (and it really looks like it, I should say), but it isn't. the information shown under the GeneView section comes from text mining pubmed articles, and for that reason you may find or not information on that section depending on the existence of literature on that particular variant. what annotation tools like ANNOVAR do is to calculate the amino acid change by placing the variant in a particular transcript (in fact, if several transcripts contain that variant position, the amino acid change would depend on the transcript and you would have to describe them all unless you precisely know which transcript does that variant correspond to), but this is different from that the GeneView section seems to be there for.