Yes, Agile in not a panacea: you can't just drop it into an existing org and expect things to be rainbows and kittens. It needs to be accepted both at the grass-roots level AND management level.
Then when Agile fails a person or an organization it's Agile's (or Scrum's) fault... when it's really a people problem.
I guess I'm just waiting for the blog articles, 10 years from now, bemoaning ROWE because it didn't save an organization with a toxic environment. :-) (And feeling sorry for the poor developers involved).
As an addendum to what Matt said, like any substantial initiative, ROWE needs support from the whole organization to succeed.
But as you mention Ryan, it is super easy for those being measured to game metrics and optimize so that they look good. Mitigating this effect is very challenging and one that we continually face but our approach, and it has held up well so far, is to focus on people over process. I.e., our results include questions like "Are developers happy?" and "Are clients happy?"
Nebulous questions with unquantifiable metrics like these keep us honest and communicative with our clients and focused on the process of setting reasonable expectations that satisfy all. We stay focused on the people and change the process to ensure results, and expectations, are met.
Results may change from developer to developer, client to client, and week to week! We expect them to change as we make progress on the project.
Then when Agile fails a person or an organization it's Agile's (or Scrum's) fault... when it's really a people problem.
I guess I'm just waiting for the blog articles, 10 years from now, bemoaning ROWE because it didn't save an organization with a toxic environment. :-) (And feeling sorry for the poor developers involved).