So, what is bad about a monolith? The complexity. Coupling of unrelated things; because it's simple to do so, because it's right there next to you and you know how it's implemented and what it performs, so you do thing A in component X knowing/relying on the fact that component Y will do the thing B, because that's how it's implemented (not because it's inherent unavoidable business logic).
How do you end up with a distributed monolith? E.g. assume that microservice B will call microservice C only after you (i.e. microservice A) had a chance to call C first. Yes your code is distributed, but you don't get the benefit of logic segregation between the modules - you just get all the downsides of distributed systems.
Do note that my scenario requires no shared code or library. Not sharing code doesn't save you from the distributed monolith - careful thought does.
All the services have their own keyspace in Cassandra, there is near-zero cross service data access