This question is difficult to answer without a definition of what is and isn't a "short-form", but here are some thoughts.
なので and だから (and けれど, が, etc.) are in modern Japanese able to function as fully independent sentence-beginning lexemes. They are not "short forms" of そうなので etc. any more than 書いた is a "short form" of 書いたり/書いたる. (However, they are obviously etymologically related to clause-final usage of なので and だから, just as 書いた is etymologically related to 書いたり/書いたる.)
As evidence, consider that as discussed in comments in many cases it is not possible to substitute a "long form" without changing the implication somewhat. Here are some intuitive observations about my own idiolect (note that I am not a native speaker).
- それだからパーティーに行かなかったの？, そうだからパーティーに行かなかったの？, etc. are not at all equivalent to だからパーティーに行かなかったの？ Note that だから is a couple hundred years old and was sometimes written with kanji making it clear that it was not considered short for anything, e.g. 然から. (That's from a mid-19th century example in 日本国語大辞典.)
- Sentences beginning with が can always be rewritten だが or ですが, but again, not それだが or そうだが in most cases.
- なので strikes me as the least resistant to a そうなので rewrite -- but still pretty resistant, and more to the point, such rewrites sound awkward and unnatural even if technically allowed.
And here are some counterarguments to the above:
- First, the prescriptive argument: なので or が have to imply something before them because なので and が can't appear independently. This is a "just because" sort of argument similar to "You have to use 'were' for the subjunctive case in English; 'If I was' is just plain wrong." I do not put much stock in this sort of thinking (although there are times when a certain strictly defined style is required, and this should not be ignored) but I include it for completeness. I do know if there is any prescriptive objection to だから or です.
- Second, the "invisible structure" argument: these words don't actually include a それ or a そう, but that at some level their use implies something equivalent to それ or そう in the structure, and therefore, the "full form" is "there" in some sense. I am not sure if any school of linguistic analysis would actually make such an argument in this case, but I include it as a possibility.