I've been trying to accomplish this via Tracing because I'm trying to figure out what's possible with it-- I've seen utilizations that suggest it can be used as a complex configurable selection tool within the Utility Network. Below I have some screenshots of the same Filter Barrier (Using a network category to select the first feature that is part of "switching" and including structures and containers) run with the Isolation Trace and the Upstream trace. The Upstream trace includes the line and the Isolation trace does not despite having the same request for each. the only difference I could find in the code (also below) when I transferred these traces to modelbuilder and then to python was "trace_type="ISOLATION"" versus "trace_type="UPSTREAM"", which suggests some back-end configuration that is not editable that causes an Upstream trace to include the line.
What I'm asking is, why are there selection differences between the two with the same starting point and the same filter barrier, and is/are there documentation or blog posts or anything dealing with complex Traces and exploring what traces are capable of? I'm looking to get as deep of an understanding as possible.