question: observability across all layers

### Feature Description

Hi team, when I'm working on dashboard for operations, I found I have to build my own observability layer (i.e., metrics/logging) around a bunch of existing util layers, for example,
- I want to know how much time is spent on concurrent limit block wait
- I want to know how many times do we retry
- I want to print out logging for rate limit
- I want to know the overall HTTP request latency with retry included (could be workaround)

I guess it's a deliberate design -- opendal takes a layering design, each layer is pluggable and doesn't interfere with other layers, but it means I need to write a bunch of observability util layers.
I'm wondering if there're better suggestions, and how other teams are dealing with the requirement? Thank you!

### Problem and Solution

Have to write customized observability layers when certain metrics are needed.

### Additional Context

_No response_

### Are you willing to contribute to the development of this feature?

- [ ] Yes, I am willing to contribute to the development of this feature.