There's a lot of research suggesting that people overestimate their productivity in remote working environments and that in general, max productivity comes from small, in-person working environments.
My guess is almost all of this is totally dependent on the environments you are discussing. Some people have built very productive home work environments but most have not. Most people's office environment is probably better than what they have at home (maybe they have multiple monitors, supplies, a nice chair and desk, etc) but I don't think there's anything specific to office vs. home aside from the investment into productivity of that space.
My guess is some of the folks on here hollering about being way better workers at home are right. They also probably spent several thousand bucks on a nice workspace, keyboard, extra monitor, stand up desk, whatever they need....
The research that I've seen supporting small in-person environments has all been carried out with low-skill workers (typically undergrad volunteers) on tasks that do not require deep critical thought or creativity. It's not at all clear that it generalizes to the sort of work that a senior technical staff member does.
I am not a researcher in the field, so its entirely possible that there is good work to support your claim, but all the research I've seen cited on HN in past discussions was crap or simply not relevant for the stated reason.
I've yet to see research that compares a modern distributed workflow with github/trac/unfuddle, IRC/campfire, IM, Skype, etc. compared to on site. Some of the remote developer studies are hilariously out dated.
Maybe there's a generation gap (People who grew up with IM/IRC can naturally use it communicate better?) or my experience is too colored by working only remote my whole career, but between the places I worked without distributed workflows and those I hear about from my friends in the industry, my anecdotal evidence says distributed teams are far far more effective.
