A decade ago we did a number of grows in the midst of an empty spare room to test about harvest mass vs light density. We used 4*1000W HPS fixtures in a closed hood and merely altered ppfd and footprint by decreasing the distance of the lamps that were aranged in an equal square. These were in the middle of the room without near walls so the plants would basically just get to intercept the direct light and not so much reflective wall-light. There was a 30% reduction in footprint, as well as less plants but an accordingly higher ppfd. Draw from the wall remained exactly the same, as did the harvest weight. This is somewhat surprising because all the science behind photosynthesis rates show that it gets less efficient the3 higher the flux is, as there is a diminishing return when gradually more and more photons do get converted to heat.
But still there are quite a number of studies arriving at the same conclusion, that is when they search for a function to describe this increase then the linear modell fits best. As long as one doesn't go overboard with it and keeps the tops still healthy. Got to add, we did prune the big fanleaves from the tops to allow the light also to reach the middle leaves, as the tops already benefit from the higher ppfd numbers. Vegging time was 2 months so there was enough foliage so that still all light got absorbed by the plants and not hit the bottom or pots. I mean, if one prunes THAT heavily then surely a reduction in yield is to be expected.
But what is responsible for this 'linear increase'?
I believe this has physiological reasons, and lies somewhere within the daily sugar demand a plant needs to grow. Some of the sugar would be used to keep the own cells alive - a "metabolic rate", which would have to be substracted from the net amount of sugar generated day by day. This is a flat/base substraction that should rise as the plant grows bigger. What's leftover can be turned into new biomass, which is what we understand or see as "normal growth". Now it stands to reason that a low influx of daily sugar will suffer more from the base substraction than a high one. To make an example:
If 50k mol (glucose) are needed for base, and 500ppfd will grant 100k whereas 1000ppfd grants a (diminished) 150k, then after subtracting the base what's left for new growth would be 50k-500ppfd vs 100k-1000ppfd which is a double increase.
Now these numbers are entirely made up just to state an example of how to explain the observation.
I'm actually looking for more scientific info into these mechanisms. That "light-compensation-point" is kind of a manifestation of such a base need where photosynthesis done at 63ppfd won't result in additional growth. Keeping mothers in a real low light environment is another experience that shows how much growth can be slowed down, with the typical adaptations of the plant to compensate the lack of sugar, e.g. only small 3-fingered leaves
