That equation was derived from data from LED lights with rows of LEDs, from 2 to 4 rows, all spaced from 2 to 3.5 inches apart. It is reasonably accurate with 40 or 60 degree optics, and with some tweaking, with no optics. But I haven't tried it with 20 degree optics.
You can design your light so that each LED lights up most of the tank, so many LEDs contribute light to every spot in the tank, or you can design it so each spot in the tank is lit by only a few LEDs. For the latter you obviously need to concentrate the light from each LED enough to get adequate PAR. There is no requirement to do it either way or a combination of the two ways - it is a choice you have to make. The more LEDs there are lighting each spot, the more you can mix in different colors without getting ugly colored shadows, and the more likely you are to have uniform lighting over most of the tank. But, this does mean the ends of the tank will be at a lower PAR than the middle, since there will be many fewer LEDs lighting the ends. If each spot is only lit by one LED it is possible to have virtually the same PAR along the entire length of the tank. But that is likely to show spotlighting effects too. It is a trade-off.
For an accurate PAR reading you have to use a PAR meter and measure it yourself. All that an equation can do is get you in the right ballpark, so a dimmable driver or drivers will let you adjust to the PAR you want.