Every diode has a sweet spot or drive current that delivers the highest umoles per watt
Yes it is less than 1% of nominal current and is typically out of the practicality of design.
your "Sweet spot" analogy suggests that there is a practical current where there is a maximum in efficacy, when in fact, efficacy continues to rise as current drops. due to thermal limitations, *every* led looks like this when efficacy is plotted vs wattage in absence of other factors
something that had a "sweet spot" (or as you previously called "optimum drive current to get the highest umoles per watt" or "maximum umoles" would look something more like this with a defined maximum current below which only incremental gains are seen as opposed to going pseudo-asymptotic on the Y-axis:
so for LEDs the "optimum" current is always the lowest you can afford to run, and is different for each user. its an economic decision (payback of efficiency/yield gains vs available capital) - there is no inherent "sweet spot" in LED current
even in that second graph we are still seeing maximum efficacy at lower current, but are experiencing diminishing returns. a true "optimum current" would generally be driven by multiple variables (like crowding of diodes to maintain a certain level of flux at canopy, and would look more like this