I have recently been given to think about certain things which Twitter is not very useful for, one of which is asking research questions to which a simple answer cannot be given. Sometimes it's just not possible to give a snap 140 character response, or refer to an online resource or even a published work. There are questions which, however straightforward they may look, need unpacking in order to tease out the problems, and the suggestion of an array of possibilities for finding out the required information.
This sort of thing was much better served by the email-based listserv, but I'm not sure that younger scholars, used to Web 2.0 social media, engage with listservs any more? I was certainly quite surprised to come across a blog post by a younger scholar talking about the benefits of forms of social media I associate with the 2010s for engaging with the wider field of the discipline, making contacts, getting one's work known about; because I found that listservs also did this, back in the Upper Palaeolithic (in internet terms) of the 1990s.
They even had archives that were searchable: though the constant recurrence of certain questions which had already been addressed on-list led me to produce the first version of the Victorian Sex Factoids Page and to compile a bibliography of recent work on Victorian Psychiatry beyond the one work published nearly 30 years that everybody still keeps citing so that I could just link to these rather than reiterating the same message.
New technology and new media do not necessarily drive out the old, which may still have their uses. The major nexus for humanities listservs, H-Net is currently undergoing a massive upgrade to bring it more into line with modern web-use practices, but there remains a value in having the potential to respond to research queries in a more discursive and nuanced fashion than 140 characters permits.