According to spec sheet, if accurate, which is no gaurantee, it produces 855 umol/s PAR.
This is enough for about 1m^2 with 12 hour operation needs (or about 150% of that for 18h operation if you exclusively grow autoflowers on 18/6 schedule). You should probably run less than 100% over 12 hours or hang at additional height above canopy in a .64m^2.
Whatever you hammer out for vege phase, multiply by 150% (assumes 18h vs 12h operation - inversely proportional to hours of use). So, if you find 50% and 18" distance works well in vege, then you'll want 75% and 18" in flower.
Best guess... you want about 60-70% in flower for an upper 30s DLI, so you need about 2/3rds of that over 18hours. 40-50% at normal hanging distances.
There is no one-size-fits-all. All recommendations requires some adjustment relative to the local variables of your garden. Temp, RH and ambient CO2 all factor into how much light per day the plant can handle.
Is the spec sheet accurate... need a diode count that mars doesn't publish anymore. If they are powering the diodes around .2 watts per, then you can rely on their spec sheet information. If it is dratically different, say .4-.5watts per diode, then you need to knock off 10-20% of that 855umol/s PAR and redo the math. Samsung tests these diodes at 0.2 watts per diode. Anything above that reduces efficacy and therefor output per watt expended.
efficacy x watts = 2.85 umol/J x 300 watts = 855 umol/s PAR
if it's 2.5 efficacy because they went cheap on diodes, now it's down to 750 umol/s PAR. You'll need another 13-15% more power (multiply by 1.15, not add .15) e.g. 60% becomes 69%. Do not assume their spec sheet is accurate. Get the diode count. Make sure it's equating to .20-.25 watts per diode and you can trust the spec sheet.
The bullshit mentioning 3.14 umol/J for the best lm301 diode effiacy is retarded, becasue it really shouldn't have a ton of 5000K diodes, which is what that value is based on. Warmer CCT is lower efficacy. 2.85umol/J is reasonable... as long as they have a proper diode count. It's a very nice light if they do.
https://led.samsung.com/lighting/mid-power-leds/3030-leds/lm301b-evo/
--- H or B does not matter. Anything that deviates from testing paramaters will reduce output per watt.