unlike hps or cmh bulbs, the efficacy is not standardized. So, depending on the efficacy of the LED you bought, it wil impact how many watts per sq ft you need.
Switch to umol/s of PAR. This is apples to apples. Daily Light Integral (DLI) is what matters. There are DLI tables to reference with simple google image search.
For ambient co2 conditions, you should shoot for 35-40DLI produced by the light at the diode. With normal hanging distances for QB and bar/strip style frames and reflective walls, it is a great starting point. No matter what suggestion you start with, you will need to observe and adjust based on plant growth. Local variables makes for different 'max' DLI in any 1 garden.
If used to metric, no conversions necessary. Umol/s of your light per m^2 and then reference the DLI table with that value and hours of use. 800-900 umol/s of par per m^2 will be roughly 35-40 dli. (2m^2? then you want the light to produce 1600-1800umol/s over 12 hours... 0.5m^2 then you want 400-450 over 12 hours.)
It's proportional to time (inversely 1:1) and umol/s (direct 1:1) output.
per sq ft..
70-80 umol/s PAR per sq ft for 12 hour operation.
67% of that for 18 hour operation.
multipling that by 10.764 will give you "ppfd," which can be referenced with hours of operation on a DLI table... you'll see it all falls in the 35-40 dLI range.
The math really is easy once you get the relationships.
The highest efficacy LED options only need about 30-32 watts/sq ft over 12 hours. (again, 67% of that for 18 hours vege or autoflowers). The shitty LED still need about 40w per sq ft for 12 hours. You can see a 25-33% difference in wattage needs simply due to efficacy and all other factors remaining the same. Efficacy is what you pay for. Higher efficacy saves electricity and reduces heat produced (usually a benefit, too). It also increased longevity. Any deviation from the diode's manufacturer spec sheet testing paramaters means shorter life / less efficacy. This is info you can look up and compare.