Firstly, all gas turbine rotors-compressor and turbine-are airfoils,
and that's why engine (and airframe) designs don't scale perfectly for
aerodynamic performance: air, in effect, has a finite size.

All engines are more efficient as the delta between the hot parts and
the cold parts increases, per Carnot. That said the primary reason car
engines run hotter than before is not because of direct thermal
efficiency but for emissions and also because smaller radiators can be
used.

Large turbines are actually designed to wear themselves in at times
as creepage brings the blades out and they ever so slightly lathe
themselves down, opening up the stator surface as they do. Active
clearance control using bleed air is another nifty feature of big
engines. Working against the economics of big fan engines is the fan
case being bigger than normal shipping means can handle. Also, a
serious FOD, midair, or controller failure resulting in an engine
writeoff is a much bigger capital hit:someone has to write a big
check.

The less the number of engines the greater probability that one
engine will fail. However the consequence of one engine failing
becomes directly higher. One engine out on a B-52 is a marginal
consideration: one engine out on a single turns your mission to worst
case recovery (find an airport, off airport landing, ditch, or
bailout/eject depending on the aircraft and terrain underneath.)

An exception to this rule is the light twin, most of which are really
1 1/2 engine airplanes to start with, and which tend to be flown by
hobbyists and part-timers. Statistics show that engine failures in
light twins kill more people than engine failures in singles, for a
lot of reasons. Cessna, Piper, and Beech knew this since roughly 1960
and their response for 25 years was, buy more liability insurance.
When high interest rates in 1986 made the reinsurance market for this
untenable, "they could no longer make reciprocating-engine aircraft".

ETOPS-Engines Turn Or People Swim!