As you point out, physical limits have a lot to do with this discussion. It is usually a big mistake to try and predict the future by following a curve and assuming it's not going to change. Somebody doing this measuring a child's growth up until age 10 could assume quite reasonably that in twenty years the kid would be nineteen feet tall.

We are not making quantum leaps like monocoque construction, monoplanes, all-metal fuselages, jet propulsion, etc. any more.

Could we send an Apollo payload to the moon with a much smaller rocket today? Modern warships actually have lower top speeds than they did in WWII (again, partly because speed doesn't matter as much).
And the modern automobile isn't fundamentally different from the cars of the 30s (no turbine engines, infinite ratio transmissions, variable stroke pistons, etc). Granted, reliability and maintenance have improved, but repairs are very difficult and performance is comparable. Even economy hasn't improved dramatically: Carnot efficiency is still the same as it was in the era of steam.

The Shuttle launch system throws a Saturn-sized payload into LEO with much smaller systems.

Modern cars are technological wonders. I can buy a non-hybrid SUV today that gets about the same mileage a VW Beetle got when I was in college.

Some of the problem may be that the advances today aren't as obvious on the surface as some of the older ones were.

It's easy to see the change from propellers to jets. Not so easy to see the astounding progress from old 707-type turbojets to the modern efficient fanjet. Look for supersonic business jets, soon. Assuming they're allowed to be built. Don't overlook the influence of modern anti-tech politics in these trends. Try to imagine building a Golden Gate Bridge today.

I understand you're talking mostly about mechanical engineering. I think we all agree that in other areas, we're just beginning.