T-lines are just a specially formed ported system essentially. There is a lot more engineering that goes into their design. Rather than reinforce only around the port tuning as a standard ported box does, a TLine reinforces a selected bandwidth. There is some ripple in this bandwidth that has to be taken into account with software. The size is highly dependent on the desired F3, you will get a slightly smaller size than a ported box for the same speaker, but still much larger than the sealed box. Certain woofers are better than others, it's a function of the QTC and Xmax from what I have picked up.
The reason they aren't more popular is because the extra calculations and effort required in designing them, and also the complexity of the build is much higher generally over a traditional sealed/ported box. You additionally have to be more careful about the high-pass on the system, similar to ported systems the Xmax of the system is rapidly approached as soon as you get below the tuning of the T-line. Most people go with active crossovers to take care of this, as well as take care of any major ripples in the line's response.
Your tweeters are all sealed-back, or else the woofer's pressure inside the box would blow out the tweeter's dome or radiating element. The independent midrange sections are required when your tweeter either is open backed or it leaks, because of this fact. Your tweeters will essentially become passive radiators to the woofers, with extremely limited x-max which will jump their gap and be ruined.
I've heard good things about T-lines response but have never personally built or heard one.
PS: On a MTM system the T-line calculations are especially difficult, because the distance from the woofer to each end of the line has to be taken into account, so each woofer's response would have to be calculated together, which can present a special set of difficulties, especially when you bring in center to center spacing between those woofers and the tweeter which you have to then shoehorn into the space remaining.
