When & why is it ok to break the equal cross section area rule?

I was watching the "sneaky heat" dvd and Erica mentioned this rule: the cross-section area should be roughly the same throughout the rmh - wood feed, burn tunnel, heat riser, donut shaped pivoting point under the barrel top, inside the barrel sides, manifold, mass tubes, chimney.

This had me thinking about some exceptions:
1) The batchbox rmh in the workshop space has an entire extra barrel above the riser.
2) The recent juicebox style chimney gets narrower than the mass tubes.

I understand how both of these choices serve a purpose (more quick radiant heat to the shop, stratified mass tubes).

What makes these work well despite its CSA not being the same throughout?

I think the most important thing is to not have constrictions. I would also say a lot depends on draft. With sufficient natural draft from the location of the RMH, there can be a lot more leeway in how the parts are sized. I would imagine though that some of the highly tuned batch burners out there need specific dimensions to work properly, or at the very least keeping all CSA's the same for everyone makes the results more repeatable and standardized.

Hi Yasmi;
Erica & Ernie built J-Tubes with a piped mass.
Batch boxes rarely utilize a piped mass but instead, more commonly use a stratification chamber (bell).
With a piped mass you want your J built to specifications, and you need a large smooth transition area where you enter the pipe system.

With a stratification chamber, things become easier.  
There is still a CSA for bell size that should be followed although, using a bypass that can be exceeded.
Your core, be it a J-Tube or a batch box needs to be built to specifications, any little changes can affect rocket performance dramatically.
Once the super-heated air leaves the riser and enters the bell, it has no choice but to rise to the top. After arriving at the top it is instantly replaced by even hotter air.  The cooler air sinks lower as the hotter air replaces it at the top.  At the bottom of your bell is the outlet chimney.  The still-hot air discovers this pipe and starts rising again(after all it's the hottest air in that pipe) this starts the draft going.

So, If building a J-Tube with a piped mass then all parameters must be strictly followed to get a successful build.
A J-Tube can be built using a stratification chamber rather than a piped mass.
Here is my shop stove as an 8" J-Tube into a brick bell.   https://permies.com/t/94980/Brick-Bell-Shop-Heater
https://permies.com/t/99519/Baby-Dragon-Roars
This design worked great until I decided I wanted to build my first batch box.
And here is where I changed it to a batch box.  https://permies.com/t/138779/Breaking-News-Shop-Dragon-admitted
And here is the latest configuration of The Shop Dragon  https://permies.com/t/181421/Batch-box-rebuild
https://permies.com/t/190599/Finishing-touches-Shop-Dragon-rebuild

Upon rewatching the video, I realize the chimney wasn't mentioned in said rule. So the solarium juicebox one is not even an exception, 8in throughout, and then there is the chimney.

Yes, I do see batch box designs having mostly brick bells. Hm. So the double stacked barrel isn't an exception either, or more like, it's a different core so the rule needn't apply.

What I'm gathering is that the J tube desires a certain something from the rest of the system (like constant CSA / chimney close to barrel) to assist rocket-y airflow, whereas in a batch box with a strong venturi effect built into the core, the core is rocket-y enough on its own to allow some design variation? Does that sound about right?

Not that I will try to innovate or anything, I will definitely stick to the Wisner J tube design when I get to build one in a backyard. I just want to feel confident that I understand the stuff presented.

Hi Yasmi,
Yes, from stem to stern, the chimney is also a part of the csa rule. As with any rule though, it can be modified if other conditions exist, like good draft.
I believe rocket scientist Glenn Herbert has such a system where his front end is 8" and the vertical chimney is 6" and says it works great.

On the other side of the coin, another general rule that Matt Walker has suggested is if you do need to have a variable system size, make the front end smaller than the back end - from feed tube to chimney exit. The proof of why this is helpful is that when you partially cover the feed tube opening with a brick, the suction (draft) increases at this point and helps to ensure no smoke back especially during the shoulder seasons where draft may be finicky.  

Peter van den Berg, has often said there should be at least a foot of space above the heat riser on a batch box in order to achieve good draft and minimise friction of the flowing gasses, along with them being much more finicky with a piped mass, hence the bell we often see on batches.

A J tube on the other hand can handle a piped mass (variable in length according to system size) but also does well with a bell.

A chimney close to the barrel is just an idea but certainly not a rule. It has its advantages and disadvantages. Good when your starting up when you need the extra draft, but not so good when your trying to save heat in your mass rather than continuously sending it up the chimney. A bypass is nice in this way in that you can control the gas flow where needed and stop it when it isn't.  

Ah right, the chimney too. Thanks for clarifying! That makes more sense as a rule.

Foot of space above the riser! I suppose it's hard to draw a single set of rules from different people innovating through different thinking.

To be fair, Peter uses a Testo gas analyzer to calibrate and modify his stoves, making his recommendations based on repeated unbiased collected data rather than just ideas or theories.

Innovation is however a wonderful thing to have fun with as it helps stretch our rules out to see what is possible.