We did a study at Media X@Stanford University when I ran the group, circa 2007, which revealed startling numbers about virtualization of work. We surveyed six large multinationals--Sun, IBM, Microsoft, Intel, HP, and Cisco--finding that, ON AVERAGE, 75% of the professional workforces in those companies worked on at least one project with colleagues on another continent; and half of them worked on three or more such projects simultaneously. Worse (or better, depending on your view), of those on such projects, 20% had NEVER met their boss face-to-face. Half of them never expected to!
How does this relate to whether you should work from home or not? Well, indirectly, it relates closely-what is the nature of the work you are doing, and with whom do you have to associate? We did a number of studies in this vein, and at Intel, the net result was that in 2008, they put a ban in place very similar to what Yahoo and HP are now doing. It cost them some of their best talent, and it is not clear from the revenue lines that it helped build more creativity, but of course the demise of the leadership PC business dwarfs all the rest of this, right?
I'd tout three books, one by my colleague Ray Price--Serial Innovators, Stanford Press, 2012, and two by Karen Sobel-Lojeski at SUNY Stonybrook. She writes eloquently about Virtual Distance, and how you build sympatico teams independently of distance or 'home schooling'.
Seems odd for a company that built the first sideband email system (on the first Cisco router sold commercially, two years before they sold their second one), that pioneered widespread video conferencing for internal usage, that pioneered HP Halo (predating and performance-wise outperforming Cisco Telepresence in many ways), that now the answer is "come to work"