The free solution was the first one worth testing. It failed, and the incomplete record changed how future surveys are logged.
If the outer positions were receiving less light, the obvious no-cost response was to raise the fixture and widen its footprint. Raising a light spreads it. That is the standard advice, and it costs nothing to try.
So I raised it and re-checked the outer positions with the same meter, at the same positions, at the same point in the photoperiod.
The outer readings decreased. The fixture went back to its previous height and has stayed there since.
That was enough to settle the decision. It was not enough to call it a documented result.
The before-and-after readings were taken as a working check and were never preserved as a standalone survey. There is no clean comparison table to publish, which means this remains an observation that changed a decision rather than a finished experiment.
It is queued to be repeated under the current logging standard, where every reading carries its position, its depth and the fixture height. The gap is the reason that standard exists.