I believe this is a problem with OKRs more then with remote work. It is too easy to game the OKR process by choosing easy but showy OKRs, doing the bare minimum and declaring victory.
I've been remote for 10 years and before that worked on a bunch of distributed research teams and it is perfectly possible to have a highly functioning distributed/remote team that really takes ownership. I mean look at open source projects.
Further office culture tends to favor a bunch of young people who all live in the same city and have time to go out for drinks after work etc. I'll take a bunch of crazy odd balls scattered across the globe doing their own thing any day.
That only kind of helps ... and can lead to micromanagement.
I've found a better approach is to focus on key metrics or KPIs and empower people to go after them without a heavy planing cycle. Like if your app is slow and buggy the OKR process tends to favor waterfally quarter long projects like "rewrite X in Y." A better approach is often to get good at monitoring and prioritize cycles spent on maintaining, optimizing, and refactoring existing stuff with a possible incremental rewrite.
I've been remote for 10 years and before that worked on a bunch of distributed research teams and it is perfectly possible to have a highly functioning distributed/remote team that really takes ownership. I mean look at open source projects.
Further office culture tends to favor a bunch of young people who all live in the same city and have time to go out for drinks after work etc. I'll take a bunch of crazy odd balls scattered across the globe doing their own thing any day.