I see that. Crawford has changed how previous editions have handled spells such as this for as long as I can remember turning this into a "fire, check target, fire again, check target" with each casting. It hasn't been that way for years. I no longer trust Crawford as a rules guy.
A new edition is going to have new rules, and new ways of doing things that are different from previous editions. That's kind of the whole point of having a new edition.
I wanted to keep house rules to a minimum, but I see that will not be possible because I have min-maxers that will exploit imbalanced rules like this cantrip that can push people up to 40 feet with no save. I won't be purchasing any more 5E books, since I don't really need them to make things in the game. This isn't the first ruling I've found to be imbalanced and changes old rulings in a bad way. I'll just go back to hoping the Pathfinder guys make a new game with simplified rules, but done with at least some semblance of verisimilitude and consistent rulings or attempts at balance which this new 5E "rules guy" seems to want to not bother with. Most rule designers understand a no save knockback over a short distance is a serious game disruption, but not Crawford. Seems to completely overlook how this ability can be exploited to trivialize the game in a bad way.
How does this break verisimilitude? Try imagining it like this: eldritch blasts are kind of like laser beams. They travel through space almost instantaneously. They fire more rapidly than bows, which is why warlocks can't move between attacks. The warlock fires a laser, it travels at the speed of
pew pew pew toward its target, and then the warlock, in the next moment, fires off another beam, and so on. I don't see that as immersion-breaking at all.
On the other hand, I think it's immersion-breaking that fighters can make 4 bow attacks in the span of like 1-2 seconds, all while running around the battlefield, and their arrows travel instantaneously to their targets over hundreds of feet, defying physics. Nerf fighters! /s
Here's a last prediction: This Repelling Blast ruling will be changed after enough people have exploited it to trivialize encounters in conjunction with ranged attacking. It'll take months, possibly years, because of how slow rule designers like Crawford and Mearls appear to be at assessing tactical rule problems without receiving player feedback. They can't seem to see the problem with their own eyes and tactical capabilities and will wait until enough people say, "It's kind of stupid that a cantrip can keep knocking any creature back 30 or 40 feet while the players move back 30 feet and keep on hitting them with ranged attacks." You may even see the truly ridiculous exploitative multiple characters with repelling blast after a two or three level warlock dip knocking a creature back double the amount to show how truly problematic this ruling is.
I don't think it's going to be changed, because it's nowhere near as powerful as you make it out to be. Repelling blast could be used to kite a single monster, assuming you can hit with enough attacks and the monster's movement speed is slow enough that it can't make up the difference. However, this tactic can easily be defeated by many, many monsters. It's only possible to kite things this way if you focus all of your attacks against a single creature, manage to hit on all or most of your attacks, and then only if the target has a fairly low speed, doesn't have some kind of ranged attack to retaliate with, and doesn't have friends that can run you down while you're doing this. That's a lot of ifs. If any of those things isn't true for an encounter, this tactic won't work. At best, you can sometimes use repelling blast to push a creature into (or off of) something dangerous, or maybe keep a single monster at bay. So what? Plenty of other classes have more powerful means of disabling or removing creatures from battle than this.
I'm going to kill it as a house rule. That's one of the nice things about having a min-max group. You get to see the problems before the general population accepts that the problem is real.
You can change whatever you want in your games, but you might want to consider trying it out in actual play first. I can tell you that it hasn't broken the games I play in yet. It's one of those things that looks much more powerful on paper than it is in practice.