OK, here goes...but from the architectural perspective...
1. Are they the right principles?
Basically, yes. But there is a some fluff and fairly obvious things in there...
2. How should they evolve over time?
Remains to be seen...you have to crawl before you walk and walk before you run...show me measurable results...ROI.
3. How do they foster the greatest amount of competition and innovation?
Huh? This is the Federal government, not the marketplace. Competition? with whom/what? Innovation? Most government institutions at all levels have lost the ability to innovate...truly innovate...its cultural...read some of the other posts here.
So now, some detailed commentary...
1. Do More with Less – Maximize ROI on IT Investments (well, yes)
- root out waste and duplication across the federal IT portfolio (agreed, but how?)
- shift to commodity IT, leverage technology, procurement, and best practices (a pop culture approach...all the buzzwords are there...)
- build on existing investments rather than re-inventing the wheel (nice idea, but still buzzwords)
Comment: "Building on existing investments" while "shifting to commodity IT" is oxymoronic. Maybe its just political pablum that non-techies can understand. Point being some legacy systems work just fine and not all workloads are suited for "commodity IT". Google's scale-out architecture works for its major function (search) in a highly parallelized manner...but I wouldn't want my bank or investment or insurance accounts running on that infrastructure.
2. Close the Productivity Gap (nice goal)
- build a “future ready” workforce equipped with the modern tools and technologies
- implement smart telework policies that give our employees increased flexibility
- reduce real estate footprint and better enable government to function during an emergency
- think strategically about how we buy, manage and use mobile devices and collaboration tools cost-effectively and securely.
Comment: Consolidation of DCs (real estate) is a good thing and I like the concept of "distributive" technologies that can better function in an emergency...all good. The rest is kinda more pablum/buzz...true productivity comes from re-engineering the processes...Business Architecture and IT Architecture...so all the talk about modern tools and mobile devices masks the real issue.
3. Improve Citizen and Business Interaction with Government
- lower the barriers to interaction with the government.
- launch a one-stop, online portal for small businesses to find and access available programs, information, and other services from across the government
- launch a dashboard where the public can track an initial set of large infrastructure projects through the review and permitting process.
Comment: Openness and Transparency are good things...gotta say this no matter what. Proof is in the pudding...what gets done and what gets opened up...
4. Enhance Cyber Security
- consolidate data centers, shut down legacy systems, and move to the cloud
- ensure Americans and our government are safe
Comment: Talking about consolidation and legacy systems here (cyber security) does not make sense unless the context is that fewer DCs should be able to be better secured with the same or less resources. Note that what some would call legacy systems (i.e. mainframes) are the most secure servers in the world (see comment about workloads above). Note also that standardization, a pre-req for cloud can promote better security...so a cloud can be more secure than the mess we have today. The second comment is more political pablum...I thought that was the mission of DHS and NSA and other overlapping functions...(see item 1 comment).
5. Change the Way We Invest
- embrace modular development
- build on open standards
- run our projects in lean startup mode.
- work with Congress to change our approach to funding technology
Comment: All good here...the key is in the funding approach, which also extends to the state and local levels, in that cities, counties and states that can do IT fairly well need to have the freedom to implement solutions (virtualization, another pre-req for cloud) and are not restricted by verbiage in bills/directives written by folks that have no clue about technology. Same goes internally within Fed IT.
Federal IT has a long way to go...buckle up for the journey. It will take years, if not decades. Luckily, demographics (aging workforce) and economics (deficit reduction efforts) will be working in our favor. I think a more realistic view is described by John Rucker here.