The people who spend time answering questions most probably already have commercial projects running.

And in a real-life situation, it takes a huge amount of information to specify a project. Even a small (tiny) project may have 50 pages of specifications. A half page of text describing how a serial channel is used (without discussing what data that is encoded into the stream) will not contain enough information to hurt a new, inventive, project.

You don't give out secrets by telling:
- how much data?
- bursty or continuous stream?
- large or small messages?
- send-and-forget or with acknowledge?
- need for error detection?
- need for error correction?
- short/long distance?
- galvanically separated?
- wire or wireless?
- low/high voltages?
- simplex/semi-duplex/duplex?
- point-to-point, bus with multiple slaves, or multi-masters?
- need for risetime-limitations?
- needs to be designed with internal hardware, or external driver chips or full controllers accepted?
- amount of code space allowed for code implementation?
- amount of RAM space allowed?
- low-volume where cost doesn't matter but quick to market?
- high-volume where development time doesn't matter, but each component counts?
- limited board space?
- fixed processors? ok to change to pin-compatible alternative? ok to switch even if not pin-compatible?
- hotplug?
- ...

A flowchart trying to compile a full selection guild would obviously quickly grow into infinity.