Precisely, I agree that this is the logical direction in which this is heading.
I think there's an implicit but rarely stated understanding - in silicon valley in particular - that there's value beyond traditional monetary capital in collecting and storing personalized information about individuals, groups, and organizations - i.e. security, profiling, recruitment and research value.
Background checks for hires, selection and identification of possible good candidates for roles, psychological profiles, etc could all - theoretically - be extracted given enough information. For now it is - allegedly - being used purely to identify and track bad elements, but these dubious 'values' exist in the data regardless (and, notably, also for attackers).
Knowing the intent of individuals and what they plan to do would clearly be massively valuable as well - PKD 'pre-crime' springs to mind, and I think Eric Schmidt was strongly signalling to the world that Google is hunting this value aggressively during this 2010 interview:
http://www.businessinsider.com/eric-schmidt-we-know-where-yo...
As connectivity spreads and devices become closer to our biological selves, this is only going to become more accurate - and thus more powerful and controlling for anyone with the ability to use and see the data.
The real questions I have are: how accurate are these predictions really (is Google anywhere near as advanced as their public statements would have us believe?), and how many organized criminals / terrorists will in their right minds continue to use these services, extrapolating, as they must, about these directions as well. With the current silicon valley mindset, the technologies to support all this infrastructure will be pursued even if for reasons of pure capital - they all align perfectly with valid use cases in the advertising, sales and marketing realms. (see: intent analysis)
I think there is a dream of purely computerized security based on 'enough' global buy-in to US-based services, aligned with sufficient communications interception. For example, if 70% of the world uses US services for communications, perhaps that is enough to identify suspicious holes/gaps/anomalies in social networks - as well as simpler patterns of criminal behaviour within the covered communications - and thus anticipate problems.
But whether this is near a reality, and whether the ones who suffer on aggregate are {citizens/residents} or {organized crime, terrorists} I think is very unclear without transparent statistics on coverage / actual crime preventions / etc. I would guess that the NSA, GCHQ, Facebook, Google et al are not there yet, but what they have done is create an arms race where if they do not follow-through, others elsewhere may do - and thus it has become a matter of national security (in terms of supremacy of the allies) after all.
Frankly I would question given the financial crisis whether they are even hunting the right targets, but direction is presumably still set from a political and defence angle as opposed to overall public good.
Given that other nations are likely now following in these same footsteps, I think it is extremely important that the US sets good precedents, because others will take their lead. And to make a parallel with the cold war, it seems like resolution of this kind of arms race would need co-ordination and agreement with other international spy agencies - after all, the volumes and value of the data/analysis they are storing is presumably as potentially destabilizing to international safety as nuclear stockpiling is.