Great Post. I would add the following to the list:

1) Be able to validate their assumptions about functions\features behaviour (which may not be sometimes clear from documentation ) by constructing quick hello world type code snippets or even using online tools.

2) Avoid Re-Inventing the wheel. Software industry has existed for a long time and has reached a stage where most of the problems we face on daily basis might have already been solved by someone else before in a better \ systems thinking way.

3) Re-Invent the wheel !! : Get a knack to understand the system you are working on, the problems and the solutions so well that you know when to re-invent the wheel.

4) Understand CPU Architectures , how the hardware really works and executes your code , Turing Machines\Von Neumann architecture, coz in the end, no matter how much software technology advances, the only real things are 0's and 1, everything else is just an opinion. ( and a Qubit in Quantum computers)