Making the trade-off

There is no best architecture.

Complexity is a cost

Microservices, event sourcing and multi-region deployment each solve a real problem and add operational weight. Adopting them before you have the problem buys the cost without the benefit.

A monolith serves most products well past the point people expect.
Split a service when teams or scaling needs genuinely diverge.

Write it down

Record the decision, the alternatives and why you chose it. In two years nobody remembers, and the record is what stops the same debate happening again.