I dont believe its nessesary to replace PBGC, more about making it more lenient, perhaps giving it a buffer window (kinda how squiglys divekicks works) to allow it to be a more consistent tool for a wider variety of the player base.
It has a pretty lenient buffer window. You just have to know when the end of your pushblock animation is up or make an effort to see it. Getting a frame-perfect PBGC reversal with some moves may be harder than others, but it's not like it's some one or two frame window.
It's not hard because of a tight mechanical check, it's just not commonly used and poorly understood. Most people that read and make an effort to learn about pushblock properties and practice it in earnest would probably have it within a day or two of practice, if that.
Doesn't that just make the game more mash friendly? Given the value you get out of a successful PBGC, I'm perfectly ok having the difficulty on it fairly high.
Different game analogy. In starcraft, the mutalisk is a fairly powerful flying unit in high numbers. They have splash damage and are quite mobile. However, their range is very short and are very weak in small numbers. The starcraft engine, being what it was, caused idle flying units to naturally spread from one another. The best players could keep the mutalisks in tight packs to get the most out of them while inferior players could reliably just point their flocks in the right direction.
It's like that with push blocks. The really good players get the most out of push blocks because guard cancelling is an option for them while slightly less good players can just use push blocks to get into a neutral state.
A) I would disagree with PBGCs being a particularly difficult technique.
B) Even if it were a difficult technique, Mike doesn't balance the game by/around execution barriers. He's done a lot to make the game accessible (look at command throws, the leniency of buffered inputs, etc.) and he's gone on record multiple times saying that if something is difficult, it doesn't have any bearing on the balance implications of that something's place in the game because he assumes players will find ways to do it consistently as they get better.
The people I've talked to that dislike it hate it because it seems somewhat random or mashy. At this point in Skullgirls' metagame we've been conditioned to appraise a pushblock on the first or second hit of a guard string or multi-hit normal as a bad thing. It doesn't push them away which people feel is the primary/sole function of a pushblock. And that makes sense. But few have paid any attention to the actual guard stun reduction properties (or hang time advantages for air pushblocks) so they get surprised by a reversal for going back in on what looked like a bad pushblock, but in reality, was a great pushblock for reasons they don't understand or care to try and understand.
This is kind of why I'm not crazy about adding a visual indicator; you had a a 25f+ visual indicator with a bright yellow flash that lets you know a pushblock took place. In my opinion good players should know what someone's options are after that, both offensively and defensively, just like wake up.
To use your analogy, throwing an overlord in your control group and move-spamming a bit wasn't the hard part, you just had to have the prior knowledge of how the heterogenous control group pathing/clumping works. From there the only thing that mattered was either how to squeeze them between mid-game turrets, or why Irradiate would be the "we late game now; you don't get to use mutas anymore" button if you were on defense. In the same vein, PBGCs (or fighting someone that PBGCs often) just take prior knowledge of pushblock mechanics.
