It's a pretty trivial problem, all in all. Modern engineering is capable of making systems that do just that on a daily basis, with a bit of predictive software. Also, that's not how nuclear weapons work in space. They do damage via X-rays, not plasma, and there isn't a hard limit on the range of their damage. But given the consistency of the damage we see in the game, they're clearly detonating at a set standoff.
The only case I know of of speed being detrimental was the AA fire control system of Bismarck, which had a minimum speed that was faster than the speed of the Swordfish. In practical terms, a faster missile is generally better. Particularly because this is Aurora, and things don't have momentum like they do in real life.
To answer the first point. A nuclear explosion in space does not, and frankly cannot produce xrays in any large number. This is because xrays are produced by atomic processes (the compton effect or bremstrallung) and in a vacuum there are no atoms to produce this effect.
That being said, it is possible to produce a nuclear device which converts some of its energy into xrays via a sacrificial "plate/sphere/etc" but this material exists for only microseconds before it is reduced to plasma by the conversion process itself. This leaves you with a warhead that is fine for frying the crap out of russian or us ICBM guidance systems but which won't make a significant impact on the material struture of the missile. The last is true so long as the range to the target is greater than a few hundred meters. Xrays like any other form of EM radiation are governed by the inverse square law.
That law tells you very clearly what range the weapon has to detonate at to be at all effective against a structural material. Xrays are absorbed inside of a few cm in a high z material and you have only microsecond or so of production so you need a lot of energy in there in the first place. 1000 m radius gives 6x10^6 m2 surface area. If I say I want a MJ per m2 energy deposition (and that isn't outrageous) then I need 10^12 joules of xrays. That is do-able with a megatonne scale nuclear device, which produces 10^15 J of energy; in general 0.1% converstion efficiency basically. But if the warhead is 2 km away then the surface area grows to 10^7 m2 and by 10 km standoff range it is 6x10^8 m2 and now you are at 100% conversion and this is principly impossible.
I have done gamma-beta coincidence measurements with nano-second timing accuracy but the system for that is entirely hardware. If I want a safety system with msec reaction times I also use hardware. A fast boolean processor with a burnt eprom program (which is nothing more than AND/OR type logic) is still 1 msec reaction times. Software is CPU cycle determined and usually is >>5-10 msecond.
For a missile moving at 50,000 km/s to arrive at <1 km from the target it needs a determination of its position relative to a moving target performed with a time acuracy of 20 microseconds and half that if the target missile is also moving at 50,000 km/s. Per microsecond the missile detonates wrongly the target is 0.5 km further away from the desired inteval. At a time error of 2 microseconds more likely then not you are outside of your engagement range.
I have no idea why you feel the bismark's guns have anything to do with missile on missile intercepts. I said that the system we use now works fine for point defence. Your chance to hit is fixed till the missile is faster than your tracking speed then it drops linearly with the target speed. Missiles that are moving too fast compared to their target have the problem of blowing past them while the missile hardware is saying "blow up" the best thing is to match speeds, maneuver close and then detonate. But for a counter missile it is nearly alway in a head to head engagement.
If the target is too slow the attacking missile can blow past it and explode too far away, if the target is too fast it can blow past the attacker and the the explosion is too far away. For soft kills likely it is better to detonate before the target and let it do the work of passing through the shell of radiation and let its guidance system get fried. My issue with the current system is that a very fast missile is always better. And it scales with speed, so a 50K km/s missile engaging a 10K km/s missile gains x5 to its accuracy. Why? The fact that one missile is faster than the other has no advantage to facilitating the intercept outside of the question of doing it before the other missile hits its target. The actual intercept is determined by how good the missile is at maneuvering into its attack position against the other missile, which is what missile agility is I would take a wild assed guess. The speed of the two missiles should not matter and if it does matter it is as said above not clearly an advantage to be so fast that time jitter makes things harder for you. Matching velocity would be best so that as far as the two missiles or missile and target are concerned they are stationary to each other.
Missile hit chance should be given by the attacking missiles agility only including anti-shipping missiles. The speed then determines only if an intercept is over all possible or how hard it is for point defence systems to engage. Speed should not be included in the tohit chance. The idea I figure was to use "target speed" to mean target agility given as a ship looses engines it becomes easier to hit. You could make it an agility to agility contest where either the missiles agilty is used or else some ratio of engine power to mass for the ship as its agility rating.
Fundamentally the current system is not consistent. If you make the argument that you are talking about standoff detonation and the detonation is a trivial matter than the missile speeds don't matter at all. A slow missiles warhead is not any different than a fast ones. Yet the slower missile is at a disadvantage even though it nearly always involved in a head to head intercept, where what matters is the closing velocity and isn't important at all which of the pair moves at what velocity. There is no mathematical difference between a 10K km/s missile intercepting a 50K km/s missile, a 50K km/s missile intercepting a 10K km/s missile and a 30K km/s missile intercepting a 30K km/s missile. Those are mathematically identical situations, but the game strongly differentinates between them. They are even more clearly identical if you say that the fuse issue is irrelevant and the standoff range issue is irrelevant. In all cases then I can switch to a frame of reference where one missile is at 0 km/s and the other is at 60K km/s and from symetry I can swap which missile is stationary and the situation is again unchanged.
I also think from a game design point of view removing speed from the to-hit chance formula would be better as it would open up the missile design possibilities more. More possibile missile designs would lead to more diversification. Unless it is just then that people min-max around agility.