I wonder whether this does not boil down to the fact that we read less than we should. We only read fresh printed articles and books because an up-to-date bibliography is something every article must have. But we probably soon forget them and we do not anyway read whatever has been written, say, 10 years ago or more and is (still) not a classic. This could amount to say that even in academic studies we risk to follow fashions as perishable as those imposing new cloths or video-games to our kids.
Moreover, we are all used to look for a word we may be interested in through powerful tools such as jstor. Do we then read the whole article? Or the whole Sanskrit passage?
Since reading takes time and is not valued as much as writing, we can only fight against this tendency through self-constraint. For instance, one could start writing a bibliographic survey of not just classics+most recent articles/books whenever considering a new field of investigation. A short summary of it could be included even in a short article.