I was talking about flywheel weight. Which we don't want to be to light.
You are right that lighter reciprocating mass. IE crankshaft, uses less energy .
Here is an article that helps explain the pros and cons a little bit.
Crank Materials and Construction
Essentially, the range of crankshaft materials runs as follows: billet steel, steel forgings, cast steel, nodular iron, malleable steel or (in some cases) cast iron. If we were to produce one crankshaft design and reproduce it in all these materials, the order of strength would approximately follow this same list. While cast cranks are typically less expensive than forgings, they can be produced in shapes not available with forgings. But dollar for dollar, forged cranks tend to be the better method of manufacture, certainly with respect to high output durability.
Often a subject of discussion and frequently believed to be critical in the design, modification and service life of a crankshaft, is how fillet radii are configured. If we were to perform a stress analysis test that included all other design features and conditions of a given crankshaft, fillet radii could be considered the most critical factor in overall design and/or modification procedure. There is belief among crankshaft manufacturers that the use of fillets of non-constant radius—sometimes called “non-circular” contours—is preferred over those of constant radius. Worst case, this is an area worth discussing with your engine builder or crankshaft manufacturer of choice.
Lightweight vs. Heavyweight Rotating Mass
Let's talk about transient torque. For purposes of this discussion, transient torque is a measure of how quickly an engine can accelerate (including under load) through its useful rpm range. Stated another way, under sudden conditions of WOT, how fast it will span from low to high rpm. From a measurement standpoint, this is torque as measured on an inertia dyno: not a so-called “accel” test as performed on an engine dyno whereby there is a controlled unloading of the power absorption unit. This is real-to-the-track” torque and it relates to an engines ability to overcome its internal resistance (inertia) to gaining rpm.
Based on these considerations, it is fair to say crankshafts don't normally operate at constant rpm. They're either accelerating or decelerating. Their resistance, in either case, includes static weight and dimensional landscape (stroke length, location and distribution of mass, etc.). Technically speaking, in a dynamic environment, crankshafts are continually changing potential energy into kinetic energy. So what, you say? Well, these are all factors that go right to the issue of how much torque is available at the output end of a crankshaft—and need to be considered for power optimization.
From a practical standpoint, acceleration of a “heavy” crankshaft absorbs more torque than one of less weight, thereby reducing the amount of net torque available to accelerate the car. But there are trade-offs in terms of durability, flexibility and potential longevity that should be considered when trimming crankshaft weight. Furthermore, it’s not all about weight. Placement of weight, relative to a cranks axis of rotation, is also important. For example, the moment of inertia (resistance to a change in state of rotation or acceleration/deceleration) increases as weight is moved away from the axis of rotation. Even between two crankshafts of the same total static weight, the one with more weight near its axis of rotation will exhibit less resistance to a change in rotational speed; it has a lower moment of inertia. Keep this in mind when adding “heavy metal” to crankshaft counterweights during the process of dynamic balancing.
Finally, where total crankshaft weight may relate to overall flexibility, it's best to err on the side of stiffness, if this can be accomplished by selecting a crank that trends toward stiffness in combination with durability and lightweight. Involved in making these type of choices, torsional vibration or deflection becomes another important issue.