"Perverse consequences" seems to understate the matter. In a footnote, Blair cites one scary example: the discovery of "an unprotected electronic backdoor into the naval broadcast communications network used to transmit launch orders by radio to the U.S. Trident deterrent submarine fleet. Unauthorized persons including terrorists might have been able to seize electronic control of shore-based radio transmitters ... and actually inject a launch order into the network. The deficiency was taken so seriously that new launch order validation protocols had to be devised, and Trident crews had to undergo special training to learn them."
Is this the only "electronic back door"? Or is it just the only one we've discovered? And if an unauthorized launch order could be insinuated into the system by hackers, why not a false-attack warning, which could generate an authorized (but mistaken) launch order? So in addition to the potential for accidental nuclear war, there is an even more disturbing threat of deliberate-but-unauthorized nuclear launches.
This cyber threat offers yet another powerful argument for Blair's de-alerting proposals.
His plan has four phases, but I'd like to focus on the first two, which seem to me to be the simplest to implement, the least likely to arouse opposition, and the most likely to reducing anxiety about nuclear "inadvertence."
In Phase 1, he recommends "revising the nuclear war plans to eliminate massive attack options and launch-on-warning from the repertoire of response options available to nuclear decision-makers." In Blair's vision, massive attacks wouldn't be impossible; they just wouldn't be an instantly implementable, one-button option in the president's "nuclear football." Neither side has an interest in deliberately annihilating the other and suffering annihilation in return. Why not acknowledge that and remove the capability from instant "inadvertent" use?
In addition, Blair suggests that "the strategic missile forces could also be de-targeted, stripped of all wartime aimpoints." At the moment, missiles are de-targeted, but—Blair says—all the old target coordinates are still programmed into computers, and missiles could be retargeted with the push of a button. In Blair's plan, retargeting the missiles would take hours rather than minutes, and missiles would be less subject to a cyber-spoofed launch order.
In Phase 2 of Blair's Oslo proposals, "strategic missiles in silos would be isolated from external launch control, by flipping a safety switch inside each silo, as was done in 1991 when former President Bush de-alerted nearly one-half of the U.S. Minuteman force almost overnight." De-alerting the rest of our missiles would prevent an electronic or physical takeover of a launch control node from causing an unauthorized launch. (That's because re-alerting the missiles would take at least 24 hours.) The Russians would simultaneously match our measures.
Blair also proposes that submarines at sea refrain from installing a crucial element of their launch system, the so-called "inverters," which would similarly preclude an inadvertent launch.
The combined effect of the two phases of his proposals, if adopted by both superpowers, would be to create a kind of time-delay firewall that would replace the short-fuse, hair-trigger, launch-on-warning, accident-prone command-and-control systems both nuclear arsenals retained from the Cold War. And the overly hasty mistaken decisions they leave us vulnerable to now.
I'll let you evaluate his third and fourth phase proposals—separating warheads from missiles and taking them out of the silos and putting them into storage—for yourself (the whole Oslo paper is here), but they would seem to require more elaborate negotiation, more commitment to complete nuclear abolition, and would likely arouse more opposition on both sides. They probably couldn't be adopted as easily as Phases 1 and 2 could be.
And there's no reason a version of them shouldn't be inserted into both parties' platforms, along with a pledge for a comprehensive study of our nuclear warning and launch procedures. It's hard to think of anything that should have a higher priority than saving the world from "inadvertently" destroying itself.