Just to chip in a small addition to this point here.
It's important to remember that they have 5 or 6 battle system designers total depending on how much Sato's role leads towards production management now.
If SE had hundreds of devs (Or even a consistent testing team of that scale with experience and skill comparable to a prog raiding static) then I don't think they would be making the significant oversights and mistakes that are pretty much a given with each expansion release. This pattern of releasing a few jobs in a state somewhere between non competitive and barely functional and then having to spend the next 6 months worth of patches simply fixing them is both predictable and a huge waste of their time.
As of right now I assume they have a similar setup to what I had at EA/Bullfrog with a large shared pool of testers and a system that simply dishes out tickets and assignments randomly as testers become available. This is backed up by how Yoshida talked about the difficulties and compromises the testing team had to make back in HW and Gordias (which in turn resulted in infamously wonky DPS checks). That's fine for a standalone single player style game or a multiplayer game that has more in common with it's peers. But FFXIV's end game is quite different to the competition and it takes time to adjust.