Excellent points regarding the tables and AI. I am a neuroscientist/cell biologist in academia and nearly everyone I know who does analysis will use MATLAB. The learning curve for Python, with its abundance of modules, object-orientedness, and lack of a standard GUI development package makes it less likely to be adopted by scientists (with the exception of those in explicitly computational fields). These seem like conveniences, but as with any software that hides the complexity of the underlying operations, a reliance on MATLAB will kneecap academics trying to transition to industry, especially because in my experience most MATLAB users do not learn about objects or even how to organize their data into tables. Python forces you to learn these skills early, which means you develop a deeper stack of data science and programming expertise. That said, once a person attains a basic competency in Python, the ML packages are unparalleled. Which again, won't matter to someone who needs to run one linear regression or classifier for the last figure of a paper. But if your bread and butter involves frequent data exploration (and explanation) using a variety of ML approaches, Mathworks has a very long way to go to come close to Python.