Thanks, I'd rather not do any digital correction in the signal path. That in my view defeats a purpose of having a SET based system with high efficiency speakers. The speakers are not dipoles. The system is in the living room combined with an open kitchen. So, there is no side wall on the right speaker side. Maybe that explains poor matching.
I'm not sure I understand. Aren't the "after" curves your response after DSP? Are you saying that although that's what DSP provides, you choose not to use it anyway? I'd also like to know how you determined the EQ you have shown. That is not a curve that will be kind on the ears.
You're quite flat between 200 Hz and 10K. There may even be little bump on there at 2K. Personally, this EQ would drive me nuts- way too bright sounding, at least for me.
As far as I can tell, playing with tubes is a waste of time if you think the choice of tube will remedy the EQ curve you have. Tube selection has far more qualitative effects than quantitative ones , and you really require the benefit of the latter with respect to frequency response more than the former. I'd be inclined to try for a more effective EQ with the Trinov, which by all accounts is a fine sounding unit. But your target curve is just not a pleasing one. Change that, and you may be well on your way to sound that makes you happy.
I'm not sure if you are familiar with the extensive research done by many (Harmon, AR, others) on what the ideal EQ curve looks like for a speaker at the listening position? This work has been performed and replicated by many but the best reference probably remains the work done by B&K and can be found here although it is over 45 years old. In brief, a flat target curve, particularly above 200 Hz, similar to the one that you have shown, is not an ideal EQ curve at the listening position. In fact, it is far from it.
https://www.bksv.com/media/doc/17-197.pdf
You might try programming your target curve to look something like the "ideal curve" shown by B&K, that looks like this:
I suggest you might be more pleased if your EQ curve looked like this, at least to the point that you might feel more comfortable playing with the EQ curve even further to achieve a curve that potentially pleases you even more. And at the minimum, i'm guessing that this exercise might be effective in convincing you that for what you seek, tube rolling is not the answer. It's really all about frequency response first and foremost. Get that right, and fine tuning with measures such as tube rolling then becomes far more rewarding and enjoyable. You've got the gear! But that EQ needs some work.
Good luck
Marty