I know that the Twitter staff already made an announcement about this issue almost a year ago, so I apologize for bringing this up once again.
The thing is that while reading the announcement it seems pretty clear that all the Twitter data, even when released for research purposes, is supposed to be made available only using Tweet IDs.
However, in section F.2a of the updated policies it is stated that:
"You may, however, provide export via non-automated means (e.g., download of spreadsheets or PDF files, or use of a “save as” button) of up to 50,000 public Tweet Objects and/or User Objects per user of your Service, per day."
My question is: what am I supposed to do, if I want to share a small dataset (about 6,000 tweets in my case) along with linguistic annotation of any kind and with a CC BY-NC-SA 4.0 license? Am I allowed to redistribute the tweets directly or should I stick to the main guideline (that is release them via their IDs)?