Government is not a business, because it has so many stakeholder needs to address; many of which are conflicting. For example, there is a whole element of public access that simply does not apply to the private sector - and conflicts between who wants access, and who doesn't. For that reason, facile claims that the federal government should operate like a private corporation fail to recognize the true scope and complexity of the challenges being faced.
Nor does NARA bashing help - to date, records management has not been a priority, and NARA's level of support and funding has reflected that. Now we have a president who has made this a priority - it is a legitimate question whether NARA has the capability to fully meet that challenge in the most effective manner. I don't think, however, that past issues necessarily should lead to forgone conclusions.
A couple of suggestions: NARA should focus on what NARA does best, and let the agencies focus on their own areas of expertise. I like the idea of a NARA run E-Records Center as a service to agencies: this creates a greater liklihood of consistency and economies of scale, and, if it can scale to the varying needs of its customer (both up and down, depending on stakeholder needs), it will encourage people to get into the system. And, such a system can be configured in such a way to avoid the cloud "boogy man".
We also need a better alignment between custodial needs and responsibilities. All too often we attempt to impose enterprise custodial duties on local custodians, who neither need the level of complexity imposed for their own business needs, nor have the skills or resources to meet those needs. NARA should focus on enterprise custodianship, and, to the greatest extent possible, give local enterprises greater flexibility in setting retentions, developing file plans, etc.... NARA then becomes the developer of the high level framework, the periodic auditor of local compliance, but leaves a lot of flexibility in the framework to meet specific local stakeholder needs.
Technology is fragmenting, developing, and expanding at a blinding pace. The policies and tools we develop will always lag behind this truth, nor should they be so onerous to stifle innovation and collaboration. By thinking in terms of a scaled framework of custodial responsibilities and needs, we can begin to align our approach in terms of technology, policy and human capital to benefit all stakeholders. People should want to practice good records management, not feel its been imposed upon them as a burden. Achieving this is the only way we can create the kind of stakeholder accountability government demands in an environment of rapid change and individual technological empowerment.