That's pretty much how I feel as well. People that complain about all the variations in CSV "formats" out there aren't wrong but in practice, for me anyway, most of the files I work with in any given situation come from the same source or a small set of sources and use the same variant.

If/when I get a data source introduced into my workflow that differs from this variant I come up with a routine to normalize it, integrate that into my workflow and move on.