I'd say that a triode with 50 watt Pda, Ra of around 1k, µ = 10 could be
made possibly using cool cathodes like the 6146, and rather smaller,
like some of the strange looking metal& ceramic transmitting tubes, but they
cost a lot more to make than a glass
bulb type.
The anode is a large external heat exchanger in some of these types of
tubes, useful where fan cooling is
wanted, ok for a transmitter where fan noise is not a problem but to make an
anode
that just needs bolting to a normal heatsink like modern transistors isn't
easy, and afaik, hasn't
been done, and couldn't be done economically, and has no need to be done
because a
glass tube gets rid of its heat through the glass radiantly quite well as
things are.

If one had a vacuum tube amp that looked identical to a transistor amp
except
for the slightly larger "strange looking" devices bolted to the heatsink,
and the giveaway OPT, would such amps have the same appeal as present
day tube amps? I doubt it.

Take a look inside a microwave oven. Its klistron(sp) is a
tube, but it doesn't look much like one. Looks like a
metal can/box. The plate was at ground potential, which
made it easy for whatever heat sinking it might have
needed, and easy for the RF coupling into the cooking area.
Its filament and grids were placed at a high negative voltage.
That's not really an option for an audio tube with OPT though.
But high voltage transistors in TV sets (like the horizontal
output transistor) are tied to grounded heatsinks in some
sets, so high voltage heat sink sil-pads exist and could
be used on tubes that would accept such. Various transmitting
tubes do look like oversized solid state devices, and are
designed to be heat sinked. They don't have that pretty orange
glow, though.....