.... If Moore's law holds with battery improvements, ....
I doesn't. It never has. It never will.
It works for computing because computing is conceptual, not physical. Concepts have no mass, displace no volume, require no energy. They're limited only by physical implementation.
Batteries are pure physical implementation. They must exist in the real world, not the cyberspace. They have to contain/deliver energy. They must be made from real materials. They must be fabricated, handled, stored and disposed of. They're limited by chemistry. The relationships of chemistry and energy are very well understood.
That's why battery evolution is always incremental. A truly revolutionary new implementation may yield a geometric improvement in the short term. And there are certainly big improvements to be made. But the long term will always be incremental and in the end they're limited by physical reality.
.... F1 has used pneumatically actuated valves for years so they can get them to operate at such high RPM without floating. Use that same technology or maybe simple electronic solenoids for the road to have an infinitely variable lift and duration for the valves based on exact engine load and rpm.
Not only would that improve efficiency, it would reduce the rotating mass of the engine by a good bit as well. Actually remove mechanical complexity in favor of simple actuators.
Don't hold your breath on that. It'll likely happen some day. Not today. Not tomorrow. Nobody knows if it's years or decades away.
Fast, reliable electromechanical actuators have been a holy grail of engineering for more than a century. As with batteries, progress is incremental. Real world physical limitations continue to be a monster to overcome.
F1 valvetrains still use cams for actuation. The gas systems replace springs. Gas has far less mass than wire springs and never fatigues or breaks. It makes sense for engines where high revs and maximum power take precedence over cost and complexity.
Electromechanical actuation is a whole 'nother ballgame.