Most powerful missiles ever made...

Are the engines gimbaled? because that is usually the problem with missing high thrust missiles I find, every instant that the gimbals do not point in the right direction with a gimbal they are still accelerating and thereby reducing your accuracy rather than increasing it, other than that I'd assume that there is enough mass to kill anything without the osmium rods being there but that's just an inefficiency and not something that would cause it to miss.

Nice, I'm glad I could help out.

the key to scoring direct hits with missiles is to lower their gimbal range to less than 5 degrees. This is because a high gimbal angle generates a sideways acceleration that the missile wants to cancel out by turning to one side.

If you draw a diagram you'll see that a desired left turn, for example, is actually accomplished by accelerating the missile to the right, where the missile will want a even bigger turn to the left to compensate, creating a feedback loop up until the missile's deviation from the ideal heading is near its gimbal angle.

The solution seems to be to make the gimbal's sideways thrust less significant compared to forwards thrust, so that thre heading adjustment for the forward movement overwhelms the heading adjustment for sideways movement.

Some one with better math is needed to calculate the ideal gimbal angle to overcome this affect I described. From experience it seems to be dependent on the acceleration of the vehicle involved, with lower gimbal angle restrictions for a higher acceleration vehicle. This is an especial problem for kinetic missiles, where the acceleration in terminal stage is often huge, a tiny angle is needed to keep them steady.

Enderminion you're doing it right when your "missile" is more massive than the fleet you're firing it at.

Owlfeathers you're right, I edited my post.

I've defied reality too, by using a fusion sport spacecraft as a missile against a fusion liner.

Crashed.

Interestingly, when I modified it to be more of a NEFP (still my favorite thing to make even though they don't truly work), it turned out to be one of the most effective on a per-missile basis that I've made. Accelerating the missile to a fast enough speed (maybe even going retrograde to the target and aiming for a high intercept relative velocity) and detonating it after enough acceleration made for a nasty impact. It had some weird effects that I'll need to screenshot for you all.

This is sort of an experiment in really fast, big missiles (110gs accel. is around about the limit before it constantly crashes the game).

As a kinetic kill around 300t:
So far, this fast missile has started combat inside the target craft.
It's gouged clean through a stock gunship and come out the other side.
Vaporized itself on a 10cm Boron armored space-station. Second hit cleanly snapped it in two. Both hits were pretty.

As a NEFP:
Gored a stock gunship with flak of various densities/sizes but at close range.
Struck the space-station and for ~20 seconds after there were constant sparks and spalling, and small cylinders (could be the flak slowed down/cooled?) were visible embedded in the armor - I believe this was the flak rebounding inside the armor, was almost always a guaranteed crew-kill.
Detonating without allowing it to dump its 6km/s delta-v in acceleration resulted in slow/weak flak showers. After it accelerates, much faster and damaging.

Implies to me that the best way to make use of a "EFP" weapon is to accelerate it as hard and fast as possible with as ~100g missile and pop it before counter-missile fire can knock it off course. The nuke won't accelerate the flak, but the missile itself can probably get stupidly fast in lieu of that. Next step is to try and make a missile with this acceleration but 10+km/s delta-v and see how hard it hits I guess.