Increasing effective engine power or reducing total weight (assuming the intent of the tech is provide significant weight savings and not, as zgrssd said, a pittance) has the same effect, except one of these options as presented has a downside while the other doesn't. Now I know the conventional small arms optimization tech says it increases ammo consumption in the tooltip but it it actually doesn't, so adding a negative downside to a linear tech would be out of place compared to the others.
Due to the way fuel consumption is calculated reducing the weight of the model by half would have pretty significant knock on effect on fuel use, so if there was a tech to reduce vehicle weight it would play double duty on reducing fuel consumption and thus competing with fuel efficiency for research priority. A 1200 weight heavy tank using a triple heavy diesel engine, without any fuel efficiency tech, consumes 34.4 oil per hex. If there was a tech to reduce total weight by half this tank would weigh only 600, allowing the design to use a double heavy diesel engine and thus reducing the fuel cost to 12.3 oil per hex. Comparatively 100% progress in fuel efficiency would reduce the original heavy tank to 11.4 fuel per hex. If both techs were applied it would cost this heavy tank 4.1 fuel per hex.
Then you also have to consider the direct relationship between weight and move modifier; a tech that reduced weight would be competing with your proposed engine power tech for improving the move modifier.
A way to increase effective engine power other than the engine design roll isn't a bad idea on its face, but having a tech that reduces ground vehicle weight steps on too many toes.