Cantonese speakers, has anyone come across this romanization before? If so, is there a name for it? The idea is that 1 and 2 tones are marked above, 4, 5 and 6 tones below, and 3 tone left unmarked.pic.twitter.com/vn0IRZrVRY
You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
I think my bias as a native Polish speaker is showing here, since my language made some stupid decisions about using a mix of diacritics / random extra letters in combination. I like the pure diacritics approach for that reason, although of course defer to people who know more.
One of the neat tricks of Jyutping is that it has unambiguous syllable separation. Once you move to pure diacritics you can get some words which are ambiguous unless properly indicated by the author. Unfortunately, the author is unlikely to notice the ambiguity.
My general opinion is that all things should be written and stored in Jyutping (as is), with representation changes done in a serialization pass just before presentation. A mechanical transformation at output time will prevent ambiguity. Of course, that requires more tooling. 
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.