Antitrust law is, I suspect,
hopelessly outdated... both the laws and the understanding of monopolies/trusts that are baked into them.
These laws were based on 19th-century competition. The problems att were price fixing, predatory pricing (eg price low to kill competition then raise prices), supply chain bottlenecking (how you gonna sell your ore without my trains) ... industrial era trust stuff.
The precedents and laws are hair-splitting and specific. It's just not the kind of system that can "think" high level and apply abstract principles to totally new problems.
Google & Facebook mostly have no prices to fix. The ad markets where they make their money are competitive bid-based, ostensibly the opposite of a "monopolistic pricing" structure.
The economic/theory just doesn't match the pratices anymore. For example: Facebooks' revenue.
Imagine that tomorrow morning BMW's revenues are cut in half. BMW would need to produce fewer cars. Cars cost X to produce. Cut X in half, and you can expect half the volume.
What would happen if we did the same to FB. My guess is that they'd still make the same FB. If you take path dependency^ out of the mix (that it's hard to fire people and adjust downward), It's scary to think how big a company is required to make FB. Doesn't seem like a stretch to speculate that it can be done on a $5-$10bn budget... 1/10th of their current revenue. After all, Facebook was Facebook on that budget not long ago.
^By path dependency I mean imagine that FB's revenue had just never gotten to $80bn in the first place, the sahare price had never gotten so high. Etc.
IDK what exactly that implies about what antitrust laws should be, but it does mean that the theoretical foundation for the current ruleset is totally off. The way monopolistic power conerts to money in 2019 is fundamentally different from 1891... I mean genuinely fundamental, I'm not using it as a superlative. The definition of monopoly, benefits of owning one, the reasons why they're bad (or not).