Honest question: what probability of finding a shooter is the “right” threshold to go in armed and expecting a shooter?
No, I wouldn't, because your facts are wrong. Yes, London police are generally unarmed, but they still get called to respond to situations where there is a likely armed suspect, and in those cases, they send in special units that are armed.
The question here -- that you avoided -- is, what is the appropriate probability before sending in such an officer, and is it higher than the 7% in the situation you cited? You were asking others to be outraged that there was "only" a 7% chance of an actual armed shooter. In an attempt to move the discussion forward, I asked what the appropriate value is, so as to appropriately judge whether such intervention was unwarranted.
To which your entire response was to shoehorn in an irrelevant factoid about London and falsely imply I'm not aware of it. (And, of course, work in a Fermi reference, because, why not?)
If I may make a suggestion: next time someone asks a substantive question, and you don't have a substantive answer, just don't reply. That's a much better approach for the forum than what you just did.
In that context, ShotSpotter's mechanism bordering on ascientific prognostication should be considered dangerous to the community because it is increasing the chances of fatal encounters for no benefit. It's basically SWATting with a fancy corporate branding.
What part of my comment is that (or anything else you've said) responding to? I asked a simple question: what is the appropriate "probability of a shooter" that justifies going in with a protocol that expects one? Now you're just guessing things that have little to do with what I said.
https://news.ycombinator.com/item?id=32259925
But yes, "tighten up" those Fermi estimates for a question no one asked and which has no bearing on the discussion.