If there are 3 different grocery stores and two of them merge, though? That's a different matter.
And if 1 of the remaining 2 is the zero-waste organic store that only rich people and hippies use? It might not even be providing all that much competition.
This implies that mergers between large services that have a network effect should always be prohibited, but why is that even a problem unless your goal is to thwart competition?
It would also create a useful incentive: Federated systems (like email) have a single network that spans entities. If Microsoft wants to buy Hotmail, they're not buying a separate network so you don't have to be worried about it even if they each have 25 million users as long as that's not too large a percentage of the billion people who use email. So then companies would want to participate in federated systems instead of creating silos like modern social networks do, because then they would be as strictly prohibited from doing mergers.
Seems like in your analogy there were plenty of grocery stores left.
People forget that the Android app and the aquisition announcement came out like one week apart. The basic web version didn't come out until half a year later.
My point is it already had a solid rich, influencer-y userbase + was about to become available on more platforms. Aquisition definitely had an impact on the user growth, but the growth itself was already inevitable.
Antitrust violations are a higher bar, you must improperly flex your dominant market position to violate the law. For that the government would have to show that FB offered an unreasonable price that nobody sane would match or that they threatened to cut off Insta links from FB if they didn't sell, something like that.