Riot is incredibly focused on "Player Experience First" - it's a core riot value and designers need to be ready to explain how their designs support it. It's brought up constantly in meetings. I love that about working here.
Fortunately, it's often possible to make designs that designs are good for both players and the sustainability of the game. That's not a conspiracy, just sustainable design. :)
The main goal of wild fragments was to give players some guaranteed fragments of their favorite champion on a regular basis, since the first "fully random" version (never intended to be final but we weren't sure what we'd supplement it with yet). This is why we were originally thinking about a "wishlist" system where players would select one or more champions to either increase their odds of rolling or guaruntee a percentage of fragments per day.
This would have worked great in theory but would be kind of annoying for players. They'd have to pre-select champions after all, and they might not understand the under-the-hood math. It would have also taken us a lot longer to build out all the extra UI and tutorials explaining it, which would have sucked because we wanted to give players a way to target their favorite champions as soon as possible.
We switched to the wild fragment solution. It was easier to explain and definitely solved the problem. The only potential downside was that it might encourage hoarding behavior. Players might suddenly feel bad about spending wild fragments in a way they hadn't about champion fragments, which would be more painful and less fun.
This was particularly relevant since we were replacing the daily bronze vault with a daily dose of wild fragments, they'd now be the major recurring reward and we wanted people to feel good about spending them. We also wanted them to spend them soon after getting them, so they got to regularly make progress on champions.
Introducing a limit accomplished both those goals. It also avoided introducing a new problem: which would be the dangers of a situation where players could "pre-finish" all future content faster than we could release it. We'd have to be super careful about giving out too many rewards in that world, which would make it harder to do experiments like putting fragments on the event passes, adding new event quests with path rewards, etc.
We were able to put those rewards in quickly, without running a bunch of economic simulations, because we knew that there was no way they could break the long-term sustainability of the mode. That makes it much easier to try stuff.
Imagine if you could "bank" XP in an MMO before an expansion. The devs would have to *reduce* XP gain to be lower overall, because it'd be kind of lame if you instantly hit level cap on day 1 of the new expansion. Level caps mean that devs can keep XP higher - because the worst that'll happen is some players might hit cap earlier than others. The new expansions will still matter to everyone.
Hoarding usually isn't great for players or devs. It means players feel worse about upgrading and makes future event quests and other activities less enticing. It also means devs have to be way more careful about giving out rewards, meaning we can't just toss them onto cool ideas; we need to spend a lot of time running careful simulations to ensure we don't create long-term problems.
The best designs are great for players and for the game's sustainability. That's always what I try for.