As far as latency issues goes - NTT is SE's chosen partner and they are the issue 100%. It's super easy to blame them, but in reality that isn't what fixes this. SE needs to work with NTT to ensure that users coming into their nodes are getting good results. Other gaming companies have done this in the past when widespread users reported issues. Before anyone tries to say it isn't widespread (because it doesn't affect them), please save it, it's documented even here on the OF in a 50+ page thread, as well as a litany of them elsewhere.

Regarding Bliztball - if it is more menu content, it will be a waste of resources/dead on arrival. The game does not need ANY more menu content. Here's the conundrum - do you design the content to be played/visited by the masses or do you simply design it to be the best it can/faithful/fun?

In my experience/opinion - these types of undertakings are best served by simply building the best possible iteration of the content form. In that, it will succeed on its own merit rather than artificially forcing it to be relevant to the masses.

What I would want out of Blitzball is a real live breathing match. Active gameplay elements, matchmaking, both solo and as a team, with ability to play against a range of AIs as well as PVP. It doesn't need to be some kind of scaling progression system, or have obscure rewards gated behind it, simply make it fun to play and people will participate.

It needs to be easily accessible, but deep at the same time. Build in functions to let FC's/statics/linkshells compete against each other for prizes. The idea is this should be something people can do during downtime for fun. Build in spectate options for people to watch.

If Chocobo Racing was basically Crash Team Racing/Mario Kart but with FF paint on it, I'd probably play it every single day with friends, but we got some awful implementation instead, basically dead on arrival.