Update 2020:

An 8 pack of 6mm thin tags that you attach to your clothes. They cost $399, and are currently sold out. They are 6.18mm thick. https://spirehealth.com

Studies show that Fitbit and Oura sleep tracking is very inaccurate. For example, the measurement of being awake agrees with polysomnography half the time. So if you are laying still but awake while in bed, fitbit will think you are asleep. I think the consumer device sleep trackers are a measure of people's willingness to believe in anything.

You can't use the results in the following quote, because the study was funded by Fitbit, but you can think of it as the best-case-scenario: “In the main group, Fitbit Charge 2™ showed 0.96 sensitivity (accuracy to detect sleep), 0.61 specificity (accuracy to detect wake), 0.81 accuracy in detecting N1+N2 sleep (“light sleep”), 0.49 accuracy in detecting N3 sleep (“deep sleep”), and 0.74 accuracy in detecting rapid-eye-movement (REM) sleep.” source

References and Reviews:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6107736
https://www.ncbi.nlm.nih.gov/pubmed/31739855
https://www.mdedge.com/chestphysician/article/193307/sleep-medicine/diagnosing-osa-polysomnography-beats-fitbit-apps
https://theskepticalcardiologist.com/2019/02/09/the-oura-ring-for-personal-sleep-analysis-lots-of-hype-and-data-little-science-utility-or-accuracy
https://www.ncbi.nlm.nih.gov/pubmed/28323455
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6025478
https://mightygadget.co.uk/rem-fit-non-wearable-sleep-monitor-review
https://www.pcmag.com/review/362712/nokia-sleep
https://www.forbes.com/sites/bradmoon/2015/11/26/review-beddit-smart-sleep-monitor/#702623303562
https://www.macrumors.com/review/apple-beddit-3-5-sleep-monitor
https://www.reddit.com/r/ouraring/comments/a6na04/how_does_the_oura_ring_sleeptracker_compare_to
https://mightygadget.co.uk/withings-sleep-tracking-mat-review-sleep-tracking-apnoea-detection-home-automation
https://the5krunner.com/2019/02/08/withings-sleep-review-beddit-emfit
https://www.nosleeplessnights.com/best-sleep-tracker
https://www.transmutableexplorations.com/oura-ring-review
https://nutritionalrevolution.org/2019/07/20/why-the-oura-ring-sucks-for-tracking-sleep-and-might-be-harming-your-health
https://medium.com/@tombarys/dreaming-about-better-sleep-dreem-oura-and-the-rest-fb29c99f0aad
https://bolesblogs.com/2019/09/20/aura-rungs-of-the-oura-ring
https://sleepshepherd.com

The best consumer review I have found of the Oura ring:
https://theunconventionalroute.com/oura-ring-review

In the comments, the author's husband Chris states: “Who knows how accurate the ring is, but hopefully it’s at least consistently wrong in the same direction every time, so it can still help us see if we’re improving…” This may be true. After all, it's measuring something, right? The studies above measured a single night's sleep of many subjects, not one subject's sleep over many nights.

My personal experience, is that even if these devices aren't accurate, they give you an idea of when you fell asleep, and when you woke up. For those that can't keep a regular sleeping schedule, that can be helpful upon waking, to know if you should keep sleeping, so your day won't be miserable. “I feel awake at the moment, but the fitbit says I only slept 5 hours. Definitely need to roll over and go back to sleep.”

I have a Fitbit Charge 2, and I find it uncomfortable to sleep with. Sometimes my hand is numb because the bulky-thick watch has been pressing against my wrist. I've used it to track my sleep for many months, and am glad I tried it out, and *feel* like I have a baseline for future measurements, however, I'm unlikely to continue to use it for sleep tracking, unless it's to compare with a sleep tracking ring I purchase in the future. I may prefer a thin watch, but no one makes a thin watch.

Update: Based on the article by Tomáš Baránek (medium.com), Dreem could be helpful in improving sleep quality. Unfortunately, it costs $500.

My experience with the Fitbit, is that it is not accurate at measuring interval training, like sprinting. It doesn't register the high heart-rates very well. I've read it has trouble taking readings if the watch is being bounced around a lot, like when you are running. In cycling I think I've had better readings (this is my personal impression, as I didn't make it a responsibility to comb through the data to give a better conclusion).

Update:

I've discontinued using the Fitbit. There's no off switch, so I can't turn it off to conserve the battery. From my understanding of lithium batteries, if a battery becomes overly discharged, it will die. So I keep recharging it. It sits on my desk.

My default goal is active minutes, rather than steps. Curiously, the fitbit quite often reaches the goal of 120 minutes of active minutes, just by sitting on my desk. When I was wearing it, is was more accurate than this, but obviously, it is optimized to be encouraging, as opposed to grading your activity accurately.

The one thing the rings can't do, compared to the Fitbit, is vibrate to give notification of a calendar event, message or call. Unfortunately, you can't set Fitbit's default handler for calls. So you will only get notifications for calls to your cellular number, and not any other number like Google Voice. Neither does it recognize Google Voice as a messaging app. Recognized messaging apps include the default SMS, Hangouts and WhatsApp.