I've never understood this concept. Reddit is the worst, though at least you can argue that they probably have all the figures to back up the idea that engagement is much higher when they push someone onto the app. For a travel app which you are only using a few times a year I can't imagine this holds.
It seems like it's product managers justifying their investment in the native applications in a lot of cases.
A company I worked for in the past explicitly made the mobile web version worse so people would use the app. They did it because with the app they could send offers through push notifications. Ugh.
It also is more likely when to users to come back after stopping using the app for a while if it's still installed and as such they can "stumble" about it.
Anyway the most well known example is probably YT which slowly reduced over time the functionality of their Web version for mobile, and also put things in to make it more annoying, too.
I don't know why so many people leave so many notifications active. I've even had people tell me that lack of notifications are the reason they don't like just using web sites on mobile.
It's less useful to focus on this single example, and more on the industry. Once these types of ideas gain momentum, they just become to be accepted as truth.
"If everybody else pushes users to their app over web, there must be good reasons to do it, we should too."
Expecting companies to rationally analyzed whether that is actually the case, is, well, expecting too much of companies.
It seems like it's product managers justifying their investment in the native applications in a lot of cases.