You should work backward from area of coverage -- are these autoflowers or photoperiods? Either? The hours of light for photos to bloom requires more power than autoflowers over 18-20 hours. it is directly proportional to hours of light per day. I'll assume 12/12 for blooming photoperiods and 18hours/day for autoflowers throughout. You need 50% more power over 12 hours to provide the same you once did over 18, or 18 with autos. So, if photos, you work backward from area relative to a much shorter lights-on cycle.
What you pay for in LED is efficacy. In general, if you pay more today, you can save money over 5-7 years of use due to lower electricity bills... also, greater longevity.
All LED dim. They dim much faster if they are run "hotter." It means what it sounds like... more watts per diode = reduced longevity. They advertise 60,000 hours of use for Samsung lm301B/H diodes, but that is only if you run them at about 0.25watts per diode. (this spec requires a percentage to make sense... 60,000 hours to 80%? 90% of original intensity? if they have integrity it is "L90" or when it reachings 90% of original)
Dividing watts by total number of diodes can inform about a lot of bad marketing -- mostly off lower-end products. You'll see more accurate spec sheets for more expensive products, but not always.
4 autoflowers -- I'd do a 4x4 tent. Get a 480watt light to cover that space for autoflowers.
If you want to do photoperiods in future, i'd get a 650w light with a dimming function. Run at roughly 80% for autoflowers.
16sq ft...
This is roughly 1.5 m^2 - makes math below easier..
So you want a minimum / maximum of: 500umol/s to 1000 umol/s -- In order to use this range you need an accurate spec sheet, FYI. The 1,000umol/s may be near a range needing extra co2 in atmosphere, but not too far over, if so. PPFD -- 500-900 (relative to 1m^2, just in case they show this instead of umol/s)
(For photoperiods: 750-1500umol/s for a 4x4 over 12 hours of light... need the above range for vege period(18h/6h --- "PPFD" would be the same for both as it is relative to 1m^2. 500-900ish, i may have used 850 in calculations -- this is old info i did the math a long time ago, but within 50ppfd is close enough)
You also want to consider footprint of light -- is it reaching sides and corners? A bar or strip-structured light is probably best option out there at moment for even coverage. Reduces hotspots of light from cob led or quantum boads... single point of light isn't optimal.. spread out is more optimal.
FC-4800 is a good quality light that will work well for 4 autoflowers in a 4x4 tent. Just an example. You can do the math on watts and diodes and see it's a 'better' light right away. This particular light would be strong enough for photoperiods. Dim to 80% for autos, but over time you'll want to slowly bump it back up to 100 over 5-7 years. This will counteract natural dimming. Think this is a roughly 0 light? might find something similar on alibaba for less than half that price but then s/h cuts into that, plus any defect or issue is a major problem to fix.
Regardless, if you want to save electricity, diminishing returns are somewhere approaching 0.25watts per diode for any samsung lm301b/h diode based board -- they'll be 90-95% of all the diodes, so we can use their spec sheet directly from samsung, which is known to be accurate. At 0.375watts per diode you still get a high quality light. When you dip to 0.50watts per diode, you are getting into the entry-level, less expensive units. Much higher than that i wouldn't touch the product with a 10-foot pole. A few hundredths higher is insignificant but efficacy diminishes at an increasing rate the further you go.
I know it is complicated... could by an entry level mars/vipar and do just as well as any high-end light. The difference will be monthly bill and longevity. In a small garden, may only be a few dollars per month saved in electricty and take 2-3 years to break-even, but then saving money after that. Regardless, the higher end lights will last another year or 2, which factors into cost of ownership too.
One other benefit (possibly a detractor) of higher efficacy is proprtionally less heat generated.. heat is wasted electricity. I have a diy light running at 0.4w/diode and a alibaba sourced light running at 0.25w/diode... the difference is insane... i even have much better heat sinks on the DiY light. The 0.25w/diode light barely feels warm, the DiY is most definitely warm to the touch, while not "hot"... both are similar efficacy due to "bin" of diode (lol, complicated AF.. bin can impact efficacy by 5-10%).
If in a hotter region, one more reason to spend a bit more money today. If in a colder region and use a space heater for your grow, you may actually benefit from a slgihtly warmer light if it eliminates need for space heater. I still wouldn't go any higher than roughly 0.5watts/diode. again, 0.53 or 0.55 is no big deal but much higher i'd just avoid it. Still have to worry about longevity a bit, here.
If you want me to look over a spec sheet and point anything out or do some number crunching, send a private message to me.
Less efficient lighting needs a bit more watts... That doesn't mean it won't grow the same quality of weed. So, feel safe about that, if you do go with an entry level product to start.
480w with the the entry level stuff (not the cheap shit, no-brand stuff on amazon) should be fine for a 4x4. after 3-5 years it may lose a step, but still put out great buds. You won't get 50,000 hours of use like you would with the FC4800 and similar quality lights. 480w with a high end light is a bit too much for autoflowers, but a dimming function or 6" more distance from canopy will take care of that. plus, you 'can' grow photoperiods of highest quality and ameliorate the dimming over a longer life of good use.