One advantage of SGML, if you are editing it manually, is that you can abbreviate tags in various ways. What exactly you can do depends on the SGML declaration for the DTD you are using. For example, in DocBook, you can write things like <literal>foo</>. In HTML, you can omit end tags, which is why you can write <p>foo ... <p>bar without closing the tags. Or you can omit tags altogether, which is why <body> is optional in HTML.
But XML was invented exactly because all these variants make parsing SGML pretty complicated. You will probably find more capable XML editors than SGML editors nowadays.
As far as processing, both forms are equivalent, because you can convert back and forth, except in some corner cases.