It's all to do with the new phones' processors. Probably in order to meet its huge production targets, Apple sourced these CPUs from two different manufacturers: Samsung and TSMC. The latter is using a 16-nanometer process, and the former a 14-nanometer process.
As a result, people running GeekBench benchmark tests have gotten two more hours of battery life out of iPhones with the TSMC-made chip as compared with iPhones running the Samsung-made chip, although they have benchmarked the same in power. Understandably, this has got people - particularly owners of the "inferior" Samsung chip-equipped model - somewhat riled up.
But Apple has issued a statement to say that it's all hunky-dory. Nothing to get stressed about. NBD, man. Etc.
The company told TechCrunch:
“Certain manufactured lab tests which run the processors with a continuous heavy workload until the battery depletes are not representative of real-world usage, since they spend an unrealistic amount of time at the highest CPU performance state. It’s a misleading way to measure real-world battery life. Our testing and customer data show the actual battery life of the iPhone 6s and iPhone 6s Plus, even taking into account variable component differences, vary within just 2-3% of each other.”
In other words, Apple's saying that unless you’re the living embodiment of GeekBench 3, which is to say you play Asphalt 8, watch videos, and use social media excessively on your new iPhone simulataneously and continuously the whole day, you should have absolutely nothing to worry about.
Benchmarking tests are designed to put unrealistic pressure on devices. But in a world where technology is so advanced that the human eye can no longer tell the difference between 401ppi and 536ppi, your own experience with your iPhone says more than numbers ever could.