My hypothesis is that few companies are big enough to be able to dedicate many people to _exclusively_ modeling. If a modeler has 4 weeks to spend on learning

My hypothesis is that few companies are big enough to be able to dedicate many people to _exclusively_ modeling. If a modeler has 4 weeks to spend on learning new things, odds are most companies would benefit more from the modeler learning how to do basic parts of the operational (python) part of their job (which needs to happen for models to be useful work) vs spending 4 weeks diving deeper on some aspect of modeling (which may or may not yield percentage points of improvement on some problem).