For example, the article mentions they knew about browsing habits but that wouldn't be the case if they couldn't see you.
I suppose it comes down to "we have the data, so let's use it"... we need to stop the mass-gathering in the first place then these "bugs" are meaningless.
Twitter runs mostly their own ads afaik, so they don't face this problem.
This is also why Twitter having problems is such a big deal; as far as I'm aware, Twitter is serving their own ad-tracking code, rather than a third-party's code. Thus, it's harder to block completely, and any vulnerabilities might (somewhat) affect even users that are running privacy extensions.
Third-party ads inserted via javascript behave as you describe.
Blocking third-party cookies (by using ublock origin or a pi-hole or firefox) helps prevent more actors from observing your online activity, but simply by being in the ad auction, third parties can get a lot of indirect information (via real-time bidding, or RTB).