That "guide" is way over generalized and simplified. It's basically like saying, "oh any 12" sub sounds good in 2 cubes tuned to 35hz and any 15" works best in 3.5 cubes at 40hz" or something like that. It also looks like he measures the line length incorrectly. The line length is measured down the middle of the line, just like measuring down the middle of the port. When it bends at 180*, you have to measure down the center but you should never measure to the wall of the enclosure like he shows.
About Sd... you don't ALWAYS have to stick with a line area equal to Sd. Also, Fi lists Sd so you don't need to measure surround to surround. Changing the line area will change the sound and it's best to use from around .5*Sd to about 2*Sd. Below that point and your output will be limited down low, there will greater velocity down the line, and there will be a spike developing in the upper bass frequencies. If you make the line area too large, you will draw line resonances in closer, have a greater chance of noticeable standing waves in some of the upper bass frequencies, and your output around the tuning frequency will be boosted. This will also cause the impedance of the system to raise. Going with just the Sd will work, but it's not necessary to follow that precisely. Then there is tapering the line which would make the line larger at the closed end than it is from the open end (or the other way around). This can help smooth out the frequency response and impedance. It will also help reduce standing waves. This requires a change in line length to keep the same tuning though.
As for choosing the tuning frequency.... you want the resonance of the driver to equal the resonance of the enclosure when they're put together as a system. This will help give you all the characteristics that you want from a T-Line. If the two resonances don't match, your line will still function but with spikes in system impedance and possibly a peaky response. You may think it's logical then to just make it the same length as the Fs. This is not the case because that is the resonance of the driver with no load on it (free-air). Contrary to the "guide" you posted, there is a load in the line that the sub sees. This load is based off of the air mass in the line. This is especially true as you place the driver further down the line. This air mass will raise the resonance of the driver. To calculate this new point of resonance, you can plot the sub in an enclosure that is sealed with the equivalent airspace in WinISD Alpha. The program will tell you the new resonance of the sub in that amount of airspace. From here, you have to play a fun game I call "match the resonance of the driver to the line length and the air mass". When you change the length, you change the air mass, which changes the resonance. You have to dial it in and there are multiple places where they meet if you change the area of your line. Another way to get the air mass you want is to use a compression chamber. You will notice that TBB uses these all the time. This can be used to lower the resonance of the driver to meet your needs for the line resonance. Changing the line area will change the air mass and you can use that to increase or decrease it.
Once this is all done though, you still need to place the driver at the correct location along the line. This is determined by the ratio of tapering on the line (if the closed end are equals the open end area, then is is 1/1=1). If you use a 1 to 1 ratio, it's usually best to put the driver a little over a third of the way down the line. This doesn't have a huge effect but it gets the resonances closer to each other and smooths out the impedance.
If you want to know more, I would highly suggest reading the derivations of T-Line design by Martin J. King. You can find them on
http://www.quarter-wave.com and
http://www.t-linespeakers.org
What I have typed here is a dumbed down way of looking at it but it's still better than the guy from your link. I hope it helps.