"Zoltan Szakaly" wrote:
The optimum thrust to weight at takeoff for a rocket is around 1.5,
and if the engines have a thrust to weight ratio of 100 then the
engine mass is 1.5 percent of the total mass at takeoff. This is more
than 20% of the dry mass.

Hellooo, McFly, I was using *actual* numbers from *actual*
rockets. If my numbers conflict with your estimates then
guess which one is most likely to be in error? The thrust
to weight ratio of the RD-180 (Atlas V first stage engine)
is about 78, and the weight is about 1% of the GLOW, yet
that's still only 24% of the first stage's mass.

Perhaps you are misestimating the dry mass fraction of
the stage.