Ha this is a really good point! I worked with so many different kinds of observability approaches and always fell back to traced logs. This might be part of the reason.