I kind of get what systemizer is saying. People may think of evolution of technologies as cycles but it is never that. A new technology 'Y' is always developed because the incumbent 'X' has some shortcomings. And even after a period of disillusionment when we revert back to 'X', it is not always the same. We synthesize the good points of 'Y' back to 'X'.

Coming to this topic, I see Microservices as a solution to the problem of Continuous Delivery which is necessary in some business models. I can't see those use cases reverting back to Monolith architecture. For such scenarios, the problems associated with Microservices are engineering challenges and not avoidable architecture choices.