In it Cody describes three key 'fumbles' which he considers an early sign of the Google's potential downfall. It's a common theme this year with many Internet prognosticators suggesting that Google's 'golden age' is starting to tarnish, with some even suggesting that it's time to find the next Google.
It's an interesting theory and one that deserves discussion, unfortunately at least one of Cody's three points is patently false. Cody's second point decries Google abandoning its much trumpeted Gmail API strategy. He makes a strong arguement, the problem is Google's never offered an API to control Gmail, let alone one which supported "hundreds of applications".
Cody's remaining points focus on content agnosticism, particularly in Google Video, and net neutrality. I agree with him that CA is going to be a big issue with Google, the recent outrage on their insertion of 'tips' in search results bears witness to that. Much has been written on the net neutrality issue, so I'll simply say I disagree with him (and agree with Google) on that one and leave it at that. But at least these two points actually exist.
To add a concern of my own, I'd say Google's lack of transparency -- particularly its poor record of communication with developers looking to use Google's systems as a back-end to their own applications -- has the potential for Google's undoing. If in years to come a competitor arrives with services as good but offering robust and transparent APIs across the board, Google could be in trouble. Of course it's not too late for them to turn it around, and their recent work on gData APIs is a step in the right direction.
UPDATE: I've had the opportunity to try and clarify the issue with Cody. He maintains that, "They most certainly did and still do offer a gmail api. But they've reduced its functionality.". When I asked for references or links he suggested I search for 'Gmail API' on Google. The results feature dozens of links to an unofficial (and unendorsed) open source user created API written by Johnvy Hwang. Cody seems to have confused this unofficial, unendorsed, and unsupported project with a real and official Google API, which it most certainly isn't. Whoops.