The only rationale given for the initial switch to microservices is this:
> Initially, when the destinations were divided into separate services, all of the code lived in one repo. A huge point of frustration was that a single broken test caused tests to fail across all destinations.
You kept breaking tests in main so you thought the solution was to revamp your entire codebase structure? Seems a bit backward.
We had a similar problem in our monolith. Team #1 works on a feature, their code breaks tests. Team #2 works on another feature, their code is OK, but they can't move forward because of the failing tests from team #1. Plus often it takes additional time to figure out if it the tests fail because of feature #1 or feature #2, and who must fix them.
We solved it by simply giving every team their own dev environment (before merging to main). So if tests break in feature #1, it doesn't break anything for feature #2 or team #2. It's all confined to their environment. It's just an additional VM + a config in CI/CD. The only downside of this is that if there are conflicts between features they won't be caught immediately (only after one of the teams finally merges to main). But in our case it wasn't a problem because different teams rarely worked on the same parts of the monorepo at the same time due to explicit code ownership.
> Initially, when the destinations were divided into separate services, all of the code lived in one repo. A huge point of frustration was that a single broken test caused tests to fail across all destinations.
You kept breaking tests in main so you thought the solution was to revamp your entire codebase structure? Seems a bit backward.