This is simpler than what I associate with mode7, because you don't need to change the view direction.
Following the second link, I can get the road drawn in perspective (Step 7), but it's adding the alternate lines in Step 10 that's confusing me.
I'm using two sprites for the road, one for the dark section with the red outline, and another for the light section with the white outline.
I can get the two colours to alternate every other line (so you get like a scanline effect), but it's the bit about scaling the sections of road so that they're thin at the top and thick at the bottom that's confusing me.
Okay, step 10 adds two blocks of code.
First block of code in there, they have an array with the z position of each line on the screen, and they scale those z positions by a constant -- this scaling just changes the thickness of the sections, you can play with the scale factor as you like.
The second block of code appears as they're looping through the lines, and it decides which type of road the current line will be. It gets the z position of the line from the zMap array (which they scaled in the first block of code). Then it says, assign which type of road it is based on that z value -- using the modulus operator so every 50 units in z it'll toggle the section it draws. Because lines at the top of the screen are 'further apart' in z, it takes less lines to move 50 units in z, so the sections are spaced closer together.
Does that make sense?