I think it's really hard to do OKRs correctly. It seems like you are working at a place that is not doing them correctly. I've also worked at places that haven't implemented OKRs correctly.
At the very least it appears that not everybody has bought into the OKRs and alignment has not been achieved.
Why don't the OKR advocates tell us how to 'correctly' implement them? Start by telling us who should define the OKRs and how granular they should be for an engineer whose priorities change week to week.
I don't think there is one way to implement these correctly. It's a people/organization problem. It's a hard to get alignment and buy in from enough people in your organization.
OKRs should at least be for a period of time of one quarter. They shouldn't be changing constantly quarter over quarter. When they do change a lot, it indicates that the company is shifting strategy. Companies need to do this from time to time, but it's not a good sign if the company is doing this every quarter.
Engineering priorities shouldn't change week to week based on OKRs, because OKRs shouldn't change that frequently.
At the very least it appears that not everybody has bought into the OKRs and alignment has not been achieved.