Not all roles are easy to measure
It's absolutely true that lazy buggers are lazy buggers in the office just like they are at home. However, there are a lot of roles that are difficult to measure. If you can measure productivity by how many lines of code someone writes or how many customer calls they field, great.
A lot of roles (probably most) have both qualitative and quantitative elements. If my team has a good throughput of tasks but pisses everyone off in the process, that's partly easy to manage and partly difficult. If they are in the office, I can see how they deal with people. If they are at home, it takes a lot more of my time to assess that and I often end up having to wait until someone complains, by which time it's already too late.
Over a long period, say a year, it's easy to tell whether someone is doing a good job and there is a limit to how long someone can hide. Howefver, carrying someone for a year is bad for their colleagues and bad for their employer and it often takes that long to find out that someone is doing some parts of his/her job badly.
There are too many people in technology that hate people and the constant toddler whining about having to do their job where they get paid to do it got old two years ago. If you all want to go to work elsewhere, good. Most of the pointy heads on the Register couldn't manage their way out of a wet paper bag but are happy to shout about how people should do a job they aren't capable of doing themselves - that's why they aren't managers.
Management should be sensible and my team does balance its time between home and the office but it's total bollocks to suggest that all jobs can be done equally well remotely and that it's simple to measure productivity.