If you take a screenshot at 5x scale, zoom in and count the number of pixels between adjacent “phosphors” of the same color, measuring from the center of the “phosphor” to the center of the closest “phosphor” of the same color:
(Options in “Dotmask” shader)
“Very compressed TV style shadow mask” - 3.00 pixels
“Aperture-grille” - 3.00 pixels
“Stretched VGA style shadow mask” - 2.00 pixels
“VGA style shadow mask” (2x2 pixel “phosphors”) - ~3.61 pixels
Which means that the “stretched VGA style shadowmask” simulation is always lower “dot pitch” and higher “TVL” than the aperture grille simulation at the same resolution. Somewhat counter-intuitive. 
(I think) The theoretical “TVL” on a 1080p display based on the above information would be:
“Very compressed TV style shadow mask”- 360 TVL
“Aperture-grille” - 360 TVL
“Stretched VGA style shadow mask” - 540 TVL
“VGA style shadow mask” - ~299 TVL
For reference, a typical 19" consumer CRT TV from the 80s or 90s had a TVL of around 300. Professional monitors (such as the Trinitron PVMs) had a TVL of around 500-600, with some reaching 700-800. At the extreme high end you have monitors like the BVMs and FW900, which are 900-1000 TVL (which is practically indistinguishable from a 1080p LCD).