Currently, security profiler displays a useful decision log which helps you figure out what's going on:

But, you as the developer are not sure:

1. where the access decision request originated, in what scope
2. what voters declared they're abstained?
3. what voters voted on what decision (assuming multiple "trees")
4. what the final decision was?

In my mind, we could reorganize this to have it be some sort of tree:

1. each decision request would be a branch with the origin class/method as the name of the branch
2. if a voter requires a decision again, we could nest it under the origin branch

Basically, the decision log should be visualized in the way in which it works (which seems like a tree) instead of a linear list.

It would even be useful if you could get this decision from elsewhere (say, a command) which could explain the decision process to you for a given set of requirements (access token, user + action etc) retroactively.