However the QUESTION is has Apple perfected the optical HR sensor the way they have done with other technologies, like Finger Print ID (though acquired tech) or Force Touch on the new MacBook?
I embarked on a crash course to learn about wrist based pulse sensors, and it's fascinating how many papers have been written on various methods / attempts to make them more accurate. It's like attempting to find the Grail of cheap sensors.
Here is some of what I've learned:
The way these things work is that they shine either red + infrared LEDs (for pulse and oximeter readings), or more recently a green LED (pulse only) into your skin.
Each LED is pulsed anywhere from 25-500 times a second, along with a dead period to check for ambient light as a base reference.
A photodiode reads the amount of light reflected back by the blood in the arteries under your skin. IR reflects most from hemoglobin with little oxygen, Red reflects most from oxygenated blood. Green doesn't go as deep and is better with darker skin and ambient sunlight IR.
The main problem with using the top of wrists is that it's not reading arteries with a true pulse (like at the end of your fingertip). Instead, it's reading from tiny capillaries, where the actual blood pulse has been dampened and spread out, which is why wrist readers are usually quite inaccurate at higher rates.
Also, movement adds difficulty. Not just mechanical movement where the sensors aren't tight against the skin, but blood movement within the body caused by the person moving their arm around or even breathing heavily.
Again, here green LEDs help somewhat because they don't read as deep. Also, offset sensors can be used to help cancel out signal noise. Which might explain the slightly odd arrangement on Apple's device.
The upshot is, everyone from major companies to students writing for a Master's degree, have tried to crack the wrist nut, and so far has succeeded only in pieces. So it'll be interesting to see if Apple's hired guns have done any better.