Which doesn't make much sense if you start making your payload increasingly expensive just to get the launch mass down.

Better design a heavy lift reusable system with an "oversized" payload capacity but relatively low launch price.

If you have to shave off mass off your payload your inevitable will add billions of costs. But if you can double the mass of your payload while keeping the requirements the same your development costs gets down fourfold.

Using nuclear engines just to make the payload a few hundred kilograms lighter would the the pinnacle of ineffectiveness. Sure, kg by kg your mission might be more "efficient" but your budget has grown exponentially.

We really have to let go of the idea that we nedd to count kilograms in our mission planning. Better just add the masses of everything you need in the end and then just launch enough propellant to get you where you want to go.

By doing this we could massively increase the science per dollar we get out of every mission.