Well we'll start with the obvious solution to the ranking system problem, and then we'll look at why they might not want to:
Problem: the problem with the current system is that it attempts to equally distribute guilds amongst the entire range. So the Top ~10% (actually more like 14% because of boundary conditions) winds up in diamond - with a big ol' cluster at 1000 (~8% of all active guilds - there's that boundary condition). The problem is that the guilds at the top end need to be differentiated in matchmaking because the fall-off in parity is *so fast* near the top. There's maybe 4 or so "real" top guilds, and maybe 20 or so guilds that can actually do something without cooperation from their opponents when faced with those top guilds. And yet we have 60+ "top" guilds tied at 1000 LP each season.
Solution: Rating decay + Slowing down movement. Basically what you need is the higher a guild is ranked, the harder it is to keep going up. This will create sparser groupings in the upper brackets and more room for guilds to fight for their position without being pushed into the punishment bracket. One way to accomplish this is to add a decay in addition to the normal rating adjustment. It wouldn't take much of a decay to adjust the distribution. With 5% decay you would in steady-state have ~20 times less 1000-guilds and about 10 times less diamond guilds. So that might already be a touch too much.
What would rating decay look like? Let's take a look at a 8 team group where everyone's rated 850:
| Initial Rating | New Rating (Current) | New Rating (With 5% Decay) |
1st | 850 | 1000 | 983 |
2nd | 850 | 975 | 933 |
3rd | 850 | 925 | 883 |
4th | 850 | 875 | 833 |
5th | 850 | 825 | 783 |
6th | 850 | 775 | 733 |
7th | 850 | 725 | 683 |
8th | 850 | 675 | 633 |
4th actually goes down a little. And none of them hit 1000. 8th actually gets kicked right out of platinum. This also means that a 1st after an 8th does not send you right back to the same spot you were at. A nice correcting influence so that you don't feel you have to *try to lose* the round after you get stuffed.
So now why would this solution potentially cause problems that they wouldn't want to just go do it:
1) People *are* going to drop leagues. Some people are going to be pissed about this as it's not a good feeling. There would be pushback on this unless somehow they managed to disguise that end result (i.e. by say "adding" extra leagues at the top and tuning the decay around the weakest diamond guilds staying steady where they're at while the strongest guilds rise)
2) The addition of decay puts more strain on the influx of points from copper guilds that can't drop below 0 (the other boundary condition) since you're taking more points out of the system. It's hard to model for those guilds at the bottom and how many of them are actually going to do something and carry some points up to be shared by other leagues. So a very conservative decay (maybe 2%?) would probably be needed to make sure the whole thing doesn't collapse and create a void between the top 3-4 guilds and everyone else (sparse good, empty not as that means unused rating space). Since decay would need to be conservative, some of the desirable effects of decay could be replaced with slower movement in general (i.e. maybe 1st/8th are +/-105 instead (+/-15 per team you beat/lose to)
3) The top group would become very static. There would be far fewer guilds bouncing in and out of it. Top guilds may get bored or exhausted depending on their relationship with the other top guilds.
TLDR: The problem with changing the ranking system is someone's not going to be happy with the changes. So the cowardly route of not changing anything is what they're opting for.