optimizing rmh in weird situation

So, asking for a friend...the rocket mass heater has only a twelve foot cob bench,, because of code reasons, and a 6" duct through that.  

It has an 8" j tube with an extra high barrel and extra large--65 gall, plus 8" off a second barrel, and that's in addition to the manifold.

But wait, there's more!  Act now and get half off your next bell.

There's a second bell for a stratification chamber (to compensate for the short bench) and this has been mostly covered with cob on the sides, about 2" .  (My friend is running out of clay but can get more in a few weeks).  The top has a tub of water in a big aluminum cauldron with a lid, maybe 20 gallons, it's squat and almost fills the whole barrel top with its girth.  It reaches temperatures of maybe 130 F after a burn.  

There's no cob hat on the first barrel but there could be...but it does have a metal "bonnet" of sheet metal curved over to reflect some heat back toward where it's needed instead of just heating the ceiling there.  (The ceiling has not been complaining about the cold).  

The bench gets only slightly warm.  The draft may still be too strong.  

The room goes back down to about 55 degrees rather quickly (ie overnight).  

The insulation situation between the first floor there and the. basement is still kind of sketchy--my friend has insulation but fitting it in around ducts and figuring out of how to remove ceiling panels has not yet occurred.

There is a crude baffle in the bottom of the exit chimney.  It could be more baffled.  (I, on the other hand, am almost fully baffled already.)

8" J tube with 6" duct shouldn't be a problem because of Charles's Law, the temperature has gone down from over 1000F to 300 by that point, and also because Uncle Mud said so.

I like to be able to cook on the top of the main barrel, including a tall pot of stew, but I'm wiling to let that go if I must, I could put a cob hat where there never was a hat.  (If you get that reference that makes me happy.)

What would you prioritize?

Do you agree that a baffle is helpful?

What else am I missing?

Thanks, team.

Bump. AI is recommending that I put an extreme baffle in, is that to be trusted?  

Another factor that I should mention is that the rocket mass heater is downstairs a two-story house, so radiant heat maybe heating up here that is gradually conveyed upstairs overnight, such that the drop-in temperature in the downstairs room is more because of that, then a lack of performance. I can try putting the thermometer upstairs for a while to see how we’re doing up there, but I have been hearing some complaints of”it’s too cold up here”

Two more wrinkles –

– The riser is more like 40 inches instead of 36. The burn tunnel was made too long, by mistake, so the builder compensated.

And then the barrel is more than 2 inches above the top of the riser, whereas Erica and Ernie’s book recommends 2 inches. That would seem to mean more turbulence and less heat transfer from the top of that barrel, so potentially more Heat would go into the stratification chamber? And then more out the chimney than we want?  The uninsulated part of the exit chimney feels like maybe 200°. The meat thermometer gives a reading up 130 Fahrenheit(but that’s not necessarily super accurate.)

Some information we need to assess the situation would be the size and character of the house, its age, insulation, weathertightness, a description of the chimney...

What exactly do you mean by a baffle? How close to the floor of the final bell barrel is the exhaust to the chimney? How tightly do you cover the J-tube feed between fires? Do you have natural draft when the heater is cold? How strong is the draft when it is operating?

As long as the main floor is not leaky (letting cold air percolate directly up), I think the underfloor insulation is not the main issue (but good to get it fully insulated). A 6" duct in the mass would possibly be too quick of a reduction from 8", though the extra large barrel(s) above the riser may shed enough heat to make that not a problem. My experience with a 6" stovepipe and chimney from the bell of an 8" J-tube is good, but that has all the heat extraction in the zero-friction bell.