I think you are close to having that Aha! moment when mapping channels make sense.
Why the UV blobs?
Each object can have multiple UV coords. For each mapping channel, it is the entire object. This is different than unwrapping with multi/sub. Since mapping channel 1 is for the tubes, they get the most UV space. However, the unwrap/mapping channel is for the whole object and the junction has to be included. Since the the junction UVs won’t be seriously used in this channel, they are dinkified and tossed aside, as it were.
The same holds true for mapping channel 2. The UV space is given to the junction, but the tubes have to be included. However, the tube UVs won’t be used in this channel and are dinkified.
Each mapping channel is an entirely different set of UV coords for the same object.
When doing this kind of thing, it’s perfectly fine to collapse the stack. For the tubes, I collapsed each time.
I selected a tube.
Gave it cylindrical (quick-n-sleazy).
Collapse.
Select next tube.
Rinse, repeat.
Once done with the tubes, I gave the junction, just the red, a quick planar and dinkified them.
Collapse.
Once happy with mapping channel 1, move on to mapping channel 2.
Same basic idea, but making sure that working in proper mapping channel when doing various things. Believe me, I’ve gone about my business and ruined a perfectly good job by working in the wrong channel. Ugh!
Collapse.
The mapping channels will remain intact and do not need to be ‘in the stack’ persay. Really up to you and how you work.
Got to take care of a few things.
Then I’ll get to uploading a file or two.