We already have the A30 mono blocks, which use 3 x EL34 per channel in SE parallel to produce 30W of power. Because this is SE and pure Class A, the transformers are heavy, about 15lb each and as a result, the amps are mono blocks. Each amp weighs about 48lb with just two transformers and a single choke.
When it comes to SE power, 30W iron will be the same size, pretty much no matter what tube is used. P-P allows for smaller output transformers to achieve the same power. I also agree that most P-P amps that I have heard do NOT sound as good as just about any SE amp, whether SET or pentode tube. However, due to the unique winding configuration of Jack's output transformers - borrowing from technology perfected in development of the A30 mono blocks - the sound is still VERY sweet and musical. It does not have the cold, hard or even diffuse sound of most P-P amps.
In my opinion, it is the use of gross amounts of Global feedback in most P-P tube amps that causes this. First of all, much of the even order harmonic distortion is cancelled in P-P config, leaving only the odd order distortion harmonics. This is why a lot of global FB is used, to reduce distortion and also improve bandwidth. Jack has discovered that IMD - Intermodulation Distortion - plays a big role in the overall sound. By focusing on ways to reduce IMD, you already get much lower THD (Total Harmonic Distortion). So, even in P-P, while there are odd order distortion harmonics present, the overall distortion spectra is very low. And, most importantly, the IMD is very low!
I first heard this when visiting Jack several years ago to learn about transformer winding. At the time, I just wanted to be able to wind my own custom value chokes for different applications. Jack taught me a lot in that time. While I was there, I heard a 10W EL34, SE amp that he had designed, based on this low IMD technology. I was absolutely impressed by the sound of this little amp. I could describe it in no different way other than to say it was special and extremely musical. We compared it to a number of different P-P and even a 300B amp that he had built and none of them had the same qualities.
Our A30 amps are built with transformers that use the same winding technique as this first 10W amp. Jack used this same technology, applied to a P-P output transformer, and found that the same results with regard to IMD could be achieved there also. When he sent me the 10W EL84 amp to listen to, I heard that same magic as the 10W SE EL34 amp I heard several years ago.
So, to answer you question, parallel SE typically does sound better. The cost is paid in the size and expense of BIG output transformers. With Jack's new P-P transformers, I am confident I can achieve the same sound in a 30W integrated with transformers sized so as to NOT give you a hernia when you pick it up.
In closing, it is a fact, I have since learned, that given enough FB (skies the limit, right?!), the output transformers can be very small and have minimal core size. You could make a 100W P-P transformer, with enough FB in the design, such that the transformer's size would be MUCH smaller than a 10W SET output. It would measure well, with good BW and low distortion. It would sound terrible however and may also have no bass. Jack has taught me that there is a true art to the design of tube amp output transformers.