In my experience, team effectiveness in different remote vs in-office scenarios plays out like this:
Most effective - All of the team physically together in the same space.
2nd most effective - All of the team remote.
Least effective - A mix of in-office and remote team members.
Most companies that try out remote work start with the least effective option (the mix). I don't know enough details of Yahoo's attempt to say if that's what they did, but I expect so.
The 100% remote option can save some office space cost, but some of that gets eaten up by increased travel. Even 100% remote teams can greatly benefit from getting together sometimes.
But one less-appreciated possibility opened up by 100% remote is the ability to draw talent from a much larger geographic area. We had engineers from all over North and South America. On balance, I think that makes it easier to retain older employees with valuable knowledge, who might want to have a house in the suburbs in a non-coastal state, instead of limiting your hiring pool to people willing to tolerate the congestion and cost of living in a big city.
