We face a reality that we have many codes to do matter modeling simulations (as described here). This is good. But, how good is it, really? For any beginner in computational modeling it is very difficult to choose a code with which to start modeling. Also, as these codes don't talk to each other, it is also hard to move between them.
If you put aside the commercial software like QuantumWise, Materials Studio, Schrodinger that are well-known materials modeling tools and are designed in a very general fashion, other open-source codes in this field more or less are designed for a specific purpose. It's true that LAMMPS, for example, is a general molecular dynamics simulation library, but most of the computational materials scientists that I know, choose their code or software based on literature review and what is the well-known application of that particular software. As I said LAMMPS, for example, is a general molecular dynamics software, but most of the people that I know which work in designing polymers/bio-polymers use GROMACS, not LAMMPS because it has suitable force-fields for that specific area of research and you don't need to reinvent the wheel from scratch.
This was just an example to show how people choose their research software and I hope it helps you.
I also remember finding it difficult to navigate the landscape of atomistic simulation engines as a Ph.D. student. While there are very extensive lists on Wikipedia, I often found it difficult to understand which of these codes actually are in broad use and for what purpose.
This was one of the motivations for me creating the atomistic.software collection, which strives to combine metadata on all major atomistic simulation engines with annual citation data from Google Scholar from 2010 until today. You can filter by license type (and more), and you can click on each citation count to see the query that was used to obtain it and browse through the articles that are citing the code.
More information in the associated preprint.