It's a nice play on the words, but there are some instructive ideas in the slide deck (available for download), if you can abide the agile-is-only-way arguments.
First, Griffiths builds the presentation around 9 bullets that are the things that PMO's are supposed to do:
Then he fills in details--from his point of view--about how traditionally managed PMO's need to change--a new game theory in his mind--to be compatible and useful to agile projects.
Of course, a balanced point of view is not Mike's forte. This is a red meat "preaching to the choir" presentation given to an agile conference of agilists.
For example, under bad old way, we read:
Monitor and control project performance – track progress against
inappropriate measures such as getting requirements fully documented and
Under agile is the silver bullet, we read:
Monitor and control project performance – track velocity, track team and
sponsor satisfaction ratings, look for dangerous velocity trends, check
backlog size, monitor iteration and release plans
Now, I count myself among practical agilists, as explained in my book, but I also know that literally billions of lines of code written the bad old way are up and working fine, so whereas I agree this generation has a good idea in agile, it's not the only game in town.
For another perspective on portfolios and agile, take a look at this:
Portfolio management and agile: a look at risk and value