So a tool that's mostly used for evil shortly after it's created is worthy of a ban (e. g. 8chan [1], Gab). But a tool which started with a more favorable ratio and slowly drifted into mostly evil-use territory (e.g. Tor, prepaid phones) isn't seen as being inherently evil and thus gets a pass. Tools that are overwhelmingly used for good will even have negative bannability (they will be actively protected and a ban will be seen as an offense) regardless of them being sometimes used for particularly evil purposes (e.g. Google, Twitter, cash).
In the present case, HKmap.live is mostly used by protesters (good usage as seen from an HN crowd PoV), so Apple's decision is seen as unjust here. But had it been mostly used by criminals early on, the consensus opinion would certainly be different.
There's a lesson for app builders: if your tool is vulnerable to bans by centralized/mob entities, it's especially important to attract enough good usage (as judged by the enforcer) early on in the history of your service.
Green Bubbles: How Apple Quietly Gets iPhone Users To Hate Android Users https://www.techdirt.com/articles/20150211/05455029985/green...
Imagine if this said:
> We have learned that an app, iMessage, has been used in ways that endanger law enforcement and residents in Hong Kong. Many concerned customers in Hong Kong have contacted us about this app and we immediately began investigating it. The app displays police locations that people send via text message and we have verified with the Hong Kong Cybersecurity and Technology Crime Bureau that the app has been used to target and ambush police, threaten public safety, and criminals have used it to victimize residents in areas where they know there is no law enforcement. This app violates our guidelines and local laws, and we have removed it from the App Store.