This is one of those things where all you will get are opinions because Google has not explicitly said whether or not it is part of their algorithm.
They have certainly hinted at user signals being part of their algorithm and engagement is certainly one of those things. How might a search engine measure those things? You might surmise that time on page would be a positive signal for a search engine to use as a measurement of engagement.
Now, some will say that Google can only determine those things if people are signed into the Chrome browser or from sites using Analytics/Search Console, and that they certainly cannot know what people are doing when using other browsers. Not sure if the other browsers matter to Google. They get plenty enough data from Chrome to assume things are roughly the same with other browsers. Also, I’m not sure how they judge things when someone searches for something, clicks a result and then searches for the same thing 20 seconds later, regardless of browser. To me, that might indicate they did not find what they wanted from the initial click.
Anyway, my OPINION, is that longer engagement on a page or site in general is the type of user signal that Google is using to judge the value of a page/site for any given search query. We have certainly noticed that creating stellar content on eCommerce websites – better than any of our competitors – has mysteriously helped with rankings – even for websites with zero backlink building to date.
As much as me, you, or anyone can look at out site log and see how many times a certain IP address visits our site, I would suggest Google is tracking this in real time as it pertains to the use of the Google search engine and the user experience back and for to the Serp, or maybe reading about X and then searching for X2 and clicking on a link and then back to Google Serp and searching for X3, and then spending an amount of time, and then comeing back to Google again and searching for Y.
I would say forget about what Google may learn from a actual visit to a site from anywhere other than from a google search. I would even suggest that data retrieved from Chrome would be secondary. But the Data from a users ( IP Adress ) behavior back and forth using Google search would be its primary data set.