If detection is mostly reliable, tracking is difficult, and the speed is high, then interception is going to be almost impossible. For ballistic missiles we're talking about M > 10 flight. For any type of gun-system, we have to detect the incoming projectile at range, track it even if it's maneuvering, and both pinpoint the location and actuate the gun system with enough accuracy to get a hit. Although high speed projectiles or lasers will reduce the need for predicting the flight-path, the real issue is still going to be tracking a high speed low-observable warhead at sufficient range to shoot it while it's in the weapon's effective envelope. That time will get shorter and shorter against faster warheads. The accuracy of the system will be hard-pushed by low-observable to generate a good track at range while the guns will be hard-pressed to aim at shorter ranges where angular velocities will be higher.
The thing that's going to allow us to get out of the stealth vs detection engineering/technology cat & mouse game is a rarely (if ever) employed implementation of radar that uses wavelengths on the order of the vehicle being searched for. Once the wavelength size is about the size of the object, nothing is invisible to radar. The trouble is that this method can't produce sufficiently accurate information for interception.
However, once the general area of the target is known, it would be easily possible to get an interception vehicle within the vicinity of the flight path. Tracking at closer ranges affords more detection options and better accuracy. If we're going to do an intercept against supersonic and hypersonic targets with low-observable technology, it's going to take everything available when it comes to accuracy. We'll have to push really push sensor designs and squeeze out a lot better kinetic performance from the rockets.
I'm expecting a combination of stealth compromising radar, less easily defeated illumination technologies, and high performance intercept vehicles with robust sensor design working in tandem with ship-based high-power painting or even coordination with UAV's.
Even if it's not missile saturation attacks from long ranges (where at least we can detect the launches due to boost motor signature) there's still the issue of the inevitable stealth aircraft penetrating defenses and launching a saturation attack at short range. The aircraft would effectively be an efficient cruise stage for a package of small missiles that only need to boost for a supersonic dash into the target. Small missiles with relatively large warheads and little warning for intercept. In terms of aircraft and missile weight (and cost) it would be very effective.
Look at it this way, we can talk all day about how to protect open-water naval ships from saturation attacks with low-observable warheads, and it will usually work out okay if we start packing enough detection and interception firepower in the correctly sized area (most anti-ship missiles designed for saturation will communicate with each other and prioritize targets), but compare how much that will cost and how much hardware we will have dedicated just to giving us the ability to sit on the surface compared to making higher value targets (ahem, giant carriers that are already nuclear powered and will always and forever more be easy to target) less vulnerable to begin with. If we can put 100m of water in between us and ASBM's, that's a lot better margin for survival than if the interception vessels take hits and suddenly it's a defenseless carrier vs. ASBM saturation attack.