However in deciding to build a SPA I failed to realize the downsides of SPAs for SEO. As an SPA darkroomfinder.org only has one route, i.e. darkroomfinder.org. Therefore GoogleBot only indexes the root domain and if someone searches for "Darkrooms" the site might surface in Google Search, but if a user searches "San Francisco Darkrooms" Google Search is much less likely to return darkroomfinder.org as nothing indicates the site has information about San Francisco.
Solving this SEO problem while maintaining the simplicity of the SPA turned out to be a challenge but I think I have arrived at a solution that checks most of the desired boxes at the cost of being tied more closely to Cloudflare's infrastructure. Cloudflare, the DNS and CDN for darkroomfinder.org, has a serverless primitive, Cloudflare Workers. These Cloudflare Workers are distributed at the edge, have KV caches, and can intercept specific paths. So my solution to having both an SPA and indexable city routes is to have Cloudflare Workers cache the city data and intercept city routes directed at darkroomfinder.org returning an HTML page with the city data and the full React SPA which boots over the HTML page. So technically it's one SPA served to many routes. GoogleBot in traversing one of these city routes, like darkroomfinder.org/san-francisco, will index the HTML page while humans using browsers receive the same HTML which the React SPA boots over.
Part of the project that remains unsolved is data collection and validation. Darkroom discovery is automated by agents running daily scans for unlisted darkrooms. But a human still has to validate that the discovered darkrooms exist and try to determine what capabilities they have such as B&W or color developing etc. I am currently working on an automated reachability and vitality checker to determine if the darkrooms already in the directory are still active, since photography darkrooms sometimes cease operations long before taking down their websites or social feeds.
darkroomfinder.org