HMMM
Ive always calculated the length of the separating port wall as the summation of the port length.
For example:
it should be calculated as (19-0.75)+(8.25+0.75), which will give you 27.25", and can be even tuned more accurately with the bend accounted for if a 45 is used. But not a big deal for a simple design.
Now regarding tuning, that......I cannot tell you unless I figure for the driver specs along with the box information together, as the driver will affect the tuning depending on the mechanical parameters of the driver and displacement volume as well. Now, with double baffle, considering dampening factors of 0.75" wood against 1.5" wood, then doubling is ALWAYS going to be a recommended idea, but for simplicity, again, usually is more of a concern with excursion and power levels beyond average. The average can be said to be the mm distance equal to the drivers sd diameter. So, if a driver has an sd of say 500 square cm, then the sd diameter is 8.8" respectively(obviously as an average, as this is just an example). So, the xmax average for that driver would be 8.8mm(just replace inches with mm).
This is a good way to figure if the driver has better than average excursion, and why some companies consider 10mm "high excursion capability" even though we may not. So, if the driver exceeds this average, it may be recommended to double baffle.
Now, for those who want to rip apart what I just mentioned, this is ONLY used as an averaged example of what I came up with for the question of double baffling. This does not mean that a driver WILL be 8.8" diameter if the area is 500 square cm, as some drivers are unique, like square drivers, or the TB 8X12. And this does not relate to the drivers sd. The sd of a driver can be calculated by measuring the ACTUAL diameter of the cone and using a formula to figure for sd from that measurement.
So, in general, if the xmax exceeds this simple little example I put together as an average, then a double baffle will be recommended...and should always be used regardless, for added support and resonance control.