Excel and GUI interfaces are fine for push-button, get-results use cases. It is always possible to build an Excel or a GUI interface on top of a Python framework like the one we present here.
However, such an interface would not be suitable to automate anything. Any automated process and/or workflow should be based directly on the Python framework, see the Entire framework in one flowchart.
The standard industry practice of setting up large spreadsheets containing hundreds or thousands (or more) of such computations in Excel, then setting up some automated process based on that is a very inefficient, error-prone approach. Formulas and dependencies are opaque, and auditing big Excel spreadsheets is a very cumbersome, error-prone task.
Here are some horror stories: http://www.eusprig.org/horror-stories.htm
This is our main use case. The users set up and run studies/tests/experiments, and produce technical documents.
Instead of setting up Excel spreadsheets in order to run computations, the users write short Python scripts. Anyone with some knowledge of Matlab (or Mathematica, VBA, etc) should be able to easily do this.
A simple Python tutorial and several example scripts should be more than enough to quickly get anyone up to speed. It is also a good way to keep acquiring Python expertise and, eventually, contribute to the Python framework.
Existing libraries with Excel interfaces can always be integrated in our Python framework as explained in Using Excel addins (third party, on Windows). This is the easy, standard, fallback solution.