It's the technological equivalent of someone shouting for help. You can assist or not.
Disclaimer: I helped build the backend so I'm biased.
Outside of that, I'm not sure what intelligence can help. Real attacks are quick, sudden. One minute I'm in traffic, next minute guy's shooting the car in front of me. And a minute later, they're gone. There's basically no value in "yelling" about these things. At best, report them on Waze so people can avoid the possible traffic.
For more drawn-out situations like domestic abuse, what's your expectation? Nearby won't intervene, but someone a few floors away is going to get an alert and what, jump in? Sure, you can invent scenarios where this might help (or even save a life) but I'd be surprised if this doesn't backfire.
People responding to crime are going to be self-selecting to be vigilantes. A vigilante guy responding to a domestic abuse situation sounds like the start of a tragic story.
Which scenarios do you see with positive outcomes? What is your team doing to mitigate the downsides?
When I was a kid (11-12) I injured my neck and was pretty worried, stuck lying down, and my pop called 911 to ask for help. A neighbor two houses down was a volunteer fireman so he heard the call over the radio. He headed over and helped make my family and I much more calm while we waited for emergency professionals to show up.
Caring about the safety of people in your neighborhood is a great start. Technology playing a role feels like less of a gamble and more of an inevitability.
Edit: A really great angle is allowing groups of people to get notified together. Works best with phone integration. At my last company I created a 9-1-1 product that does exactly that. A user could load neighbours or relatives to be alerted and listen in to their 9-1-1 calls. That way if, say, an elderly relative called 9-1-1, nearby family could perhaps intervene (for minor stuff or cases where the caller is suffering from a mental lapse) or at least know what's going on immediately.
This is also a great value add to businesses or campuses that want to know about emergencies in real time for coordinated response. Or just to know when paramedics come in the front door, who is having the incident so they can efficiently direct them.
I'm a trained medical first responder (lapsed cert.), so I definitely can appreciate what you've stated. In fact, I'd probably react the same way as your neighbour. I'd pop out and help my fellow neighbours too.
When you present it that way, I like the idea. I'm not sure about the way Vigilante is presented though.
If you don't actually have a clear answer to that, then this is a half-baked idea because it's not clear what problem you're really solving – and it falls back to the idea of "ZOMG! AWARENESS!!!!1" which is really a sign of insufficient thinking, IMHO.
I'd challenge you all to step back, articulate a clear problem frame and goal and maybe then look at the way crime unfolds – map out before, during after and look at what awareness does at each of those points (and what signals you can meaningfully find). How, then, can you build an app to do those things/support those aims and activities that you've identified as your goals? If you've done that, then articulate that in your description/writing.