Light spreads out according to inverse sqare law -- the gist is that it spreads out SUPER FAST.
e.g. if the light source is 1 foot away from canopy, then another foot deeper will be 1/4th of that intensity. this is based on a single point of light and would apply accurately for an hps bulb. LED didoes are numerous points of light and that does mitigate this effect to some extent, but it still spreads out very fast. The point is maximizing light hitting the top of the canopy is far more prodcutive than lower levels per square inch of leaf.
Based on the 1-foot example, it should be obvious that stronger light from further away will retain more light intensity further away from light source, i.e. greater penetration, but there's a point at which the extra electricity to do so is not worth the minimal benefit.
Some backgound biology that is relevant (not an opinion, like the 1-foot example above):
The products of photosynthesis are highly mobile. It really doesn't matter where it is produced. It freely moves around vascular tissue due to concentration gradients - as it gets used, it creates a "pull" in the vascular tissue to even out concentration across membranes. Those temporary concentration gradients are caused by apical dominance dictating where resources are distributed. No matter where the light hits, the dominant growth will get most of the resources. Will local leaves provide a greater proportion of the sugar used in nearby bud development? Sure, but that's more about path of least resistance than anything else. The plant is still giving the lion's share of resources to the dominant growth regardless of this fact, and sugars from non-lcoal leaves still flow freely and ever-striving for even concentration throughout the vascular tissue. (overall health and optimal levels of sugar present being a different conversation - for this conversation, these factors are assumed to be competent or already ideal for simplicity)
anecdote -- just because so many repeat this nonsense, I have purposely covered up bud sites with leaves for the last couple years. Yet to be able to notice any difference in density, leaf ratio or quality when comparing those buds to buds of similar dominance that were exposed to more light. So, whatever affect these people think they observe it must be incredibly tiny, if it is even true, which i doubt at this point. this is not proof, but if there was a correlation, you'd expect to be able to observe it, lol. And if it's too small to see, who gives a fuck anyway?
One of the most important things a canopy can do is absorb all the light -- no holes for photons to pass through and completely miss the plant. LEaves closer to the light get significantly stronger light, so maximizing penetration to leaves further away is mathematically not a good idea for the most part. Maximizing light that hits the closest leaves to the light is a far better strategy, far more efficient capture of energy.
Removing leaves for airflow might be good, but only due to a self-inflicted problem of over-crowding your canopy in the first place. Sticking to ~3 colas per sq ft, give or take, will automatically give good airflow and 'enough' light penetration without any additional effort. You'll still max out yeild per sq ft given ambient co2, which tells you there are enough colas that you aren't missing out on any yield. You simply won't do much better than 50-70g/sq ft with ambient co2. If reaching those levels, that means you are capturing all the light that the plant is capable of using per 24hours and therefore fully making use of the CO2, which is the limiting factor. Proof is in the pudding.. otherwise you'd get greater grams per sq ft production with better light penetration.
Sumary:
Leaves intake co2 - don't remove them, lol. CO2 is the limiting factor. Apical dominance is what dictates where resources are used not where the light hits the plant Concentration gradients push the flow of sugars around the plant. Cause and effect... not opinions. this is what you should extrapolate from when choosing what to do or believe.
----
Saying DLI isn't useful is like saying you don't need a measuring tape... You could just eyeball the distance you hang your lights, right? Then, through trial and error each time you hang it figure out the proper distance that causes optimal growth patterns. but, isn't measuring it, taking note of that distance, then starting from that point a hell of a lot easier? This may still require a minor amount of trial and error / adjustment, but significantly less. DLI is simply a measurement. You either now how to use it or you don't and then get all insecure about its existence, lol.
for some reason it's gotten incredibly popular to believe that education and generational knowledge is somehow not useful, bwahah. That just reaks of insecurty and a fragile ego.
DLI allows for an apples to apples understanding of light intensity applied to the canopy regardless of hours of operation or size of garden. It is not necessary for this conversation because this conversation is more about proportional intensity and not the exact values in this case. It is assumed you are providing max DLI relative to your local variables -- and "your" maximum will be different from someone else's due to varying CO2 levels, temperatures, and RH%. Something you have to observe and react to the plant whether you have a quantum meter or not. A plant that resides in a piss-poor VPD can not use as much light as a plant in a more ideal VPD environment. But, if you know you should be around 35-40dli, it can help you start at a point which requires less trial and error to find what your local environment allows as far as light intensity provided.
You don't "need" a measuring tape to build a garage, but it is damn useful, lol. You could cut and recut lumber until you get the structure square, but why not measure twice and cut once? DLI is similar. Ignoring it only makes for an extended trial and error process. Ignore it if you want to expend twice the effort to do the same basic things, bwahaha.