One fundamental reason we get overwhelmed by our incoming information and the tasks we need to do is not clearly defining the difference between an inbox and a backlog.

They are both effectively lists, so there is nothing to distinguish them on the surface, and you could treat one as the other relatively quickly. But this is the way to madness and a feeling of there being too much.

So what is the difference?

An inbox is a raw capture that requires processing. An important thing to note is that there is a decision to be made with it by processing. That decision could, of course, be very systematic and, therefore, automated.

A backlog's not about decisions but performing a task. A decision is a task, but the distinction here is that there are no decisions about the item itself. Tasks and other items (even inbox items) can be generated by going through a backlog, but in itself, it isn't something to be decided at a meta-level.

To give some guidance on the distinctions and examples, the obvious one for inboxes is email. But another one may be capturing ideas. You should have some inbox to capture ideas (we often have thoughts and ideas that don't necessarily fit into a predefined area, and it adds friction to organise when capturing). These then need processing in some way, e.g. are they tasks or reference material or something to read later.

Two of the three examples above are backlogs: tasks and read later. We don't need to decide on a task. We only need to perform it (even if the task is to choose something). Likewise, with a reading list, we only need to read it. This is why they are backlogs and not in boxes.