Long story short -- if longevity of the bridgelux diodes is similar to teh samsung, then you might as well flip a coin. Technically, you will provide more umol/s at 480w with the FC4800, but it is a 19umol/s difference... which adds up to nothingn you will be able to discern at home.... maybe, in a 1000 plant crop that could be different? at some point scale might make it noticeable, but it'd be a very large scale to do so.
---------------
efficacy, or umol/j is more about cost savings and longevity of the unit. it does not impact quality or spectrum of light provided, but is one of the largest factors in regard to cost of the product... much more than any variation of 'spectrum'... though some things impact it more than others. for example, uv led diodes will drag down the efficacy value. The cool white will be more efficient than warm white. reds will drop it too.
you'd hve to delve into the spec sheets for samsung and bridglux diodes. i am quite familiar with lm301 specs, but not so much with bridgelux.
i know a well-treated samsung diode will last 50-60k hours before it is reduced to 90% of orignal intensity... (these longevity curves increase in loss over time, so it only goes faster from there... but you certainly don't have to throw out a light with 90% of it's original efficacy, either... but also wouldn't delay too long.) As a bloom only light, it could last 10 years if the driver holds up before you reach 50-60k hours. then you got some period of time beyond that, if needed.
so, there are number of photon/s produced and with various proportions of RBG etc etc... for lack of proper term i'm going to refer to that as your 'spectrum' (not quite right, but close enough....)
Then there is the efficacy of producing those photons, which is expressed as umol/J... These are different concepts. (w/s = j/s, this is why it remains 1:1 with umol/w math.. it's not quite right but same math so who cares in this context... just divide umol/w, which is an equivalent numerical value to umol/j)
Spectrum .... Efficacy... evaluate each on their own. you can have a perfect spectrum and a low efficacy and still grow amazing plants... it'll just cost more kw-hours to do so. Quality of buds will not be different if the spectrum and all other factors are the same.
We'd need more decimals to do this accurately -- the 2.7 and 2.8 aren't precise enough... (2.7666 vs 2.80625, difference = .0396 umol/J, so even smaller than it looks -- 1.5% difference in output, same watts)
one light is producing a few more umol/s tan the other at same wattage... but, probably won't see a difference in outcomes with the naked eye or any number of plants commonly found in a non-commercial garden.
so, compare the longevity curves.. read into bridgelux diodes... as long as their longevity is similar, there is virtually no meaningful difference between the two. i assume it's the same heatsink and such? that would be a factor to consider too.
The differing diode count is likely no big deal.. the fc4800 is running the chips at an average of less than .25w/diode, which will match up with the specs on samsung.com. Looks like they run fewer watts through the bridgelux, so make sure the spec sheet there is using the same watts/amps per diode or the info needs to be translated a bit. running hotter only reduces efficacy, but that will be relative to each brand and not something you can apply to different brands blindly.
you can find a similar question about CBX3590 cree cob LED vs Luminus cxm32??? on youtube (i forget the exact luminus model#)... BUT... Day 1 out of the box, the luminus outperforms the cree cob led... however, after just 10,000-12,000 hours of use the luminus cob starts to show its wrinkles... while the cree cob led keeps going strong for another 40,000 hours or more before it decrapitaties to the same point (80 or 90% of original, i don't recall if L80 or L90 used). The longevity curve for the luminus cob led sucked... sucked bad... sucked eggs. "those" luminus cob led suck compared to the high-end cree cob even though their spec sheet on the surface looked better.
*not a parallel here about samsung quality... sammy chips are good. and, i'd feel fairly confident bridgelux makes a decent diode, but you gotta do your due dillegence. find a longevity table and compare. Merely that you need to look beyondt he basics to see any real difference.