I'm doing my first SSB homebrew project and am learning a lot as I go.
It's for 60 meters.

I've done the speech amp, DBM, crystal filter, transmit mixer, and one
stage of post mixer amplification. This has taken me to about +6 dBm,
and I want to get to +30 dBm. (I've already built a 1 watt in / 10
watt out "final" amplifier.)

My question has mostly to do with the number of stages and operating
class. As I approach 1 watt, class A gets pretty impractical due to
device dissipation, or so it seems. But I want to carry waveform
faithfulness as far as possible. Is it accurate to say that if a class
that doesn't give full swing in both directions is used, a harmonic
filter should be used before the next stage? I'm trying to infer from
what I see in the literature. I'm generally seeing a single device
(not push pull) in AB (I think) just before the output filter, but not
often at the lower levels.

I guess a push-pull stage could be used to give waveform linearity to
get to the 1 watt level. I just don't know if that's needless
complexity(?)

The next decision is whether to go from +6 to +30 dBm in one stage or
not. I'm inclined to think 14 dB in one stage followed by 10 dB in the
next would be best. Any guidance?

73--Nick, WA5BDU