Wouldn't you also need to include things like, time from founding, revenue per employee, growth rate, return on investment, etc. to evaluate whether Atlassian or Trello has been the more productive business?
The GP's point still stands: closed offices _may_ make for more productive employees (for some classes of employees), but it is not obvious that they translate into a more productive _business_.
Instead of using Trello, you could compare Atlassian against Fog Creek. Fog Creek was founded two years before Atlassian, and both companies are more or less in the same space - they make productivity tools for developers. And yet it's hard not to look at Atlassian as the more successful business so far.
