Your first paragraph's implied question and your formal question are different.
The first concerns an already finished work, which is just unpublished, and the second a work in progress. I'll comment on the second one as the more interesting one to discuss.
When you make your work in progress public, you just let people see your ideas and use them or comment on them. I perceive a publication of unfinished work as an invitation to take a look and try to answer questions like
1) What else can be done this way?
2) Can you do it better (find some shortcuts, generalize, whatever)?
3) Do you see how to finish the project or why it is a dead end?
4) Are there any mistakes in what is done so far?
The advantage is, of course, getting a few more brains engaged, which has the potential of enhancing the quality of the final product drastically. The "disadvantage"
is that you agree by default to share the credit with anyone who contributes something of value. Which one is more important for you is up to you to decide.
Outright scooping is not something I would be particularly afraid of; not in my field at least (my usual mode of operation is to send what I know to a few "experts" and ask what they think rather than to put unfinished works on arXiv but that is primarily because I do not want to clutter my arXiv list of papers with half-cooked stuff some of which will never be finished and some of which makes little sense). Most people who get interested in your project share their thoughts with you on exactly the same basis you share yours with them: if you go open and supply your ideas for free, they pay back with the same coin.