chore: allow automatically replacing a node even if it is active as an API BN

[r-birkner]

When a node is redeployed, but hasn't been removed from the registry before that, it is automatically replaced: the old node record is removed and the new one created.

Initially, this was only possible for unassigned nodes. Then, functionality was added such that this also worked for nodes assigned to a subnet.

This PR enables the same behavior for API boundary nodes. When a node that is currently active as API boundary node is redeployed, the old node record and the old API boundary node record are removed, and a new node record and a new API boundary node record are created.

We can't just use any node as API boundary node. The node needs to be configured with a domain name. This case will fail as we have an invariant in invariants/api_boundary_node.rs that ensures that each API boundary node record has an associated node record with a domain name.

In addition, I removed the test should_replace_node_in_subnet in do_remove_node_directly.rs as it is a literal copy of should_replace_node_in_subnet_and_update_allowance with one difference in the comment of line 458 and 528.

[github-actions]

If this pull request affects the behavior of any canister owned by
the Governance team, remember to update the corresponding
unreleased_changes.md file(s).

To acknowldge this reminder (and unblock the PR), dismiss this
code review by going to the bottom of the pull request page, and
supply one of the following reasons:

1. Done.

2. No canister behavior changes.

Done.