Both pokemon are very popular in NU; Liepard's win condition is to kill Sawk while leaving Golem alive, then set up Stealth Rock. Liepard cannot use Assist much before Sawk dies or the win condition is ruined. Liepard will commonly lose; if you're an experienced player with a little creativity, you can switch in Gothorita to trick a Sticky Barb to Sawk and then stall it out. This takes foresight, skill, and critical team building.
Oh, you are going to switch to Sawk on my Ditto? Let me just Nasty Plot up here and either phaze you back out or stall your Close Combats with Substitute until you get weakened enough for +2 Dark Pulse to OHKO. This can also appliy to Samurott who is phazed back out after Nasty Plot and risks a lot by switching back in if the SRer is phazed in again. AssistPard can adapt to counter-situations just as well, or even better, than Riolu can (if only because they are much rarer).
Kangaskhan needs to die or Liepard needs room to sub in order to make the win condition consistent; you may still win, but bad luck with Whirlwind will have Liepard dead before you can deal substantial damage to your opponent's team. In this scenario, you can use Ditto to deal with Kanga or you will have to get thinking on another way to wear it down; your team has little offense aside from Ditto, Focus Punch being unusable against a standard Kangaskhan. You can win, but you need effort and a little elbow grease.
Know what I use to deal with a Kangaskhan? Riolu, who can tempt Kangaskhan to go for Fake Out + Double-Edge (which Riolu survives) then proceed to Counter for a OHKO after the latter. Alternatively, forcing her to spam Double-Edge to weaken herself or Tricking her are also options.
Say your opponent is using RestTalk Cradily. Unless you have a nasty plot set up, Liepard will usually lose the 1v1 against Cradily. Your win condition is to find room to plot, sub, or both while whirlwinding through your opponent's team.
Or you could Trick the Cradily to render it useless. I will agree it is a handy stop to ShufflePard, but this isn't like Jynx, which practically forces you to run a check/counter but at least there are multiple, nor is Jynx a completely unstoppable force otherwise. Cradily is your only stop against ShufflePard in most cases, while your other members are useless against it. Trick Gothorita could very well screw it over too.
So;
"Uncompetitive" is a poor reason to ban something.
The Liepard assist strategy is not skilless; it simply has an alternative win condition and good and bad team matchups.
List the amount of bad team matchups that aren't singular Pokemon. Only a handful of Pokemon can actually stop the strategy, and if you don't have these the match could be decided turn 1.
If you look at metagame balance with your mindset, the game gets overcentralized, predictable, and boring. Your post does not tell us why the strategy isn't fit for competitive play, it tells us why the strategy is good.
A good strategy is one thing, but an overcentralising one is something else. The whole meta could otherwise delve into "use ShufflePard or use ShufflePard counter", there won't be any sort of diversity as a result. We'd essentially be seeing either Liepards or Cradily/Swellow/misc...if you don't have any of these, you won't survive long in the meta. Even then, ShufflePard has access to a slew of techs (the aforementioned Riolu and Gothorita) that could swing the matchup in its favor.