Why not instead auction them off or maybe add a business model to cut some losses? It can't hurt more than shutting it down completely?
My guess is big companies get some sort of tax benefit from doing this? But is that the sole reason behind it? Or am I missing some part of the picture?
If you has to focus in order to success, you are already on the edge of fail. Remember Gmail? It was a 20% time casual product.
And what was the first thing Larry Page did when he took over as Google CEO? Start hacking away at projects that were going nowhere to focus on a smaller number of key areas.
The other thing is most of the people who developed this product have left Google - so maintaining a rapidly ageing code base is tough...
Mixpanel, sosh, quixey, pathjoy, formula am, etc