Yes, the advertised cycle life is under a specific test, which is not likely to be an exact fit for many use cases, but is probably not wildly unrealistic for some uses. I think solar garden lights (if they have conservative low voltage protection) and LED candles could potentially be similar.
For a standard Eneloop, it would be a 500mA discharge rate, which should be in the 50 lumen range for many flashlights, and in the range I’ve seen some AA-powered motorized toys draw. Each test cycle drains the cell by about 60% of nominal capacity. Looking at the most recent Eneloop test I see on HKJ’s website, the voltage under load at that drain and state of charge was about 1.25V. Some finicky devices stop working at that voltage*. Or some users may charge pre-emptively before it is empty to reduce the likelihood of running out while using a battery. Charging for the test cycle is done at a 4 hour pace, with theoretically a bit of over-charge.
Capacity is tested every 50 cycles. Keep doing this until capacity is down to 75% of nominal, or voltage during cycles 1-49 drops below 1.0V.
But it is definitely common to discharge batteries deeper than 40% state of charge, and to use high charge and discharge rates, all of which accelerate the wear on the battery. Also, the IEC test does not monitor internal impedance as a life cycle criteria, so you can have batteries that fail to perform in high drain devices, even though their capacity under slow discharge might seem ok.
AA Cycler runs his tests down to 0.9V, and he charges them at twice the IEC rate. Those differences compared to the Eneloop rating per the IEC test standard gives a hint of how cycle life can vary depending on use.
Keep in mind, only 300-400 cycles on an Eneloop is still a lot of use. If you charged once a week, it’s 6 years.
* I still use some alkaline batteries in low value devices, mostly kids toys. I haven’t bought alkalines in years. When I get low on alkalines, I check the battery disposal bin at work. Probably 1/4 or more of the alkalines I find in there are over 1.4V, which is about 75% full. I assume this is because of these devices that don’t work well at low voltages, combined with the voltage drop of alkalines under load, or people discarding batteries pre-emptively.