ABM unreliable, require multiple interceptors per incoming - exchange favors attackers who can always saturate. Also expensive, interceptors usually cost close or more than what they're designed intercept. And currently for anyone not US, just getting US to divert 100s of billions to homeland ABM defense is a win as it will take resources from other priorities. This is also assuming ABM works against hypersonics, including all the recent talk about flak/cloud defense. It's purely speculative.
> blow up one target once
For capital assets that takes billions and years to replace like carriers, subs, supply ships, long range bombers, awacs, critical infra nodes all you need to do is blow it up once. It doesn't take a huge pool of hypersonics to essentially dismantle major US force projection capabilities.
> you just won a nuclear response
Why? Primary motive is deterrence, but with conventional forces. And if it comes down to actually exchanging conventional hits, it will follow escalation ladder, no need to cripple, or cripple to point of existential threat that warrants moving to nuclear.
Who cares if you can blow up one target once? Unless you marshal enough to wipeout enough infrastructure to really cripple your opponent, it won't do you much good anyway; and if you do cripple them, and they're nuclear, congratulations; you just won a nuclear response. You now have bigger problems.