30-40% from a normal distance (16-18" is as close as i'd get with a quantum board design, but with 95w might be able to get closer than that)... distance from canopy should be about good coverage without too much sacrifice in middle (i.e. the geometry of it). 100w can cover less than 3 sq ft on a 12/12 light schedule. If using photoperiod plants, this is what you want to work backward from. With an autoflower on 18/6 you can cover 150% more area. it is proportional.
(read up on Daily Light Integral - DLI - don't need to know the math just understand it's rate of photons multipled by hours of use and relative to 1m^2 - makes appples to apples comparison. if i give 35 DLI it's the same as a different sized garden on a difference time cycle giving 35 DLI)
100-300 ppfd is all you need. MAx - 300 / 10.76 x 3 = 84 umol/s -- I can't find specs for that light, but probably 240 ish max. 33% should be fine. it'll be lower than the calculation due to some portion of light not actually hitting the garden area or absorbed into wall.. that's fine.
This is meant to give a rough ballpark. you then observe and react to the plants. take notes at what keeps them from stretching early. That's what you want to do next time. the calculation just gets you within a few inches or a few percent. I see this light only has 4 dimming options, so 25% and a couple inches closer is doable with tiny plants in a tiny space.
as sthey stretch, you meet those new needs with more power... quickly ramping up by 2nd week, i bet.
the advertised coverage seems accurate -- 1.5x1.5 for photoperiods in flower 2x2 for anything on an 18/6 cycle. i was a bit too optimistic above with my guesstimation, but not far off, either.