This is because there is no use-case for professional users.

When it comes to computational chemistry there are two possibilities:

The computational load is low, but also accuracy like the implemented force-field method. Now you can calculate molecular properties so fast that it feels interactive.

Or you use more accurate methods (hybrid density functionals). For decently sized molecules (>100 atoms) you now might want to consider using HPC (speak supercomputer). Now you have something which is state of the art, but it's not interactive anymore.

If you however use computationally cheap methods, you probably want to calculate properties of a LOT of molecules (high throughput calculations, screening). This means you must either use known structures, or generate them programmatically (using some genetic algorithm usually). Now you don't need anything interactive either.

I don't know, it can still useful for generating your initial cells for complicated systems (defects, interfaces, etc.), and it can always be useful to test an idea qualitatively before you do a high-accuracy calculation.

There is so much software because there are many different needs, each with specialist requirements. Most likely you want to look at PyMol. However, even though that's one of the more approachable packages it will require some domain knowledge.

It gives you access to a lot more properties of the molecule you build than just the energy: Thermodynamics, solvation energies, vibrational frequencies etc.