having worked on products where analytics matter, it's important to be specific about behavior of those who use your products, and a better metric would be related to lingering, did you stop scrolling and if there are many tweets in this feed where you stopped, did you move your mouse towards one of those tweets? was your lingering intentional or related to inattention? It's important to develop an understanding of which in the list is the one you are lingering on (or what even is that list). It's important for behavior monitoring but also to make sure you are serving ads in a way that aligns with attention (assuming you run an ad supported platform, like Twitter) and that you can have accurate measurements for ad impressions (scrolling past an add without even stopping shouldn't cost an advertiser more than the base cost of the ad, and should cost more if someone lingers for even a half second.)
Other random ideas for what you would consider counted? you replied, you liked, you clicked a link through the tweet, you expanded a shortened version of a long tweet. There are so many ways to determine "read" and assuming it's always the naive "it was served" is like building a business based off of page views and not actual engagement.
It's impossible to know for real if someone has read a tweet -especially on Twitter where several tweets appear on the screen at once - however "dwell time" is a common measurement that is usually enough to appease marketers.
In this case it doesn't matter though because the aim is to mitigate cost, so equating an impression with engagement is fine.
