Each fan header can have different settings. If you daisy chain 3 fans off a single header, those 3 fans will operate identically. Most fan hubs operate as a single fan header too. In your BIOS, you can define the source (typically CPU or Mobo) and a preset fan curve (typically Quiet, Standard, Performance, and Full). These presets adjust the fan curve, which is a mapping of source temperature to fan RPM. The Quiet preset would typically require higher temps before ramping fan speeds up, while Performance will do that at lower temps. You can also define your own fan curves.
Better mobos will all have fan ramp up/down (hysteresis). This is how fast the fans increase/decrease their RPM. This is a key setting in dealing with fan noise. Instant ramp up/down means that when the sources reaches a threshold, say 50C, the fan RPM might go from 25% to 60% instantly. Add a few seconds of the ramp up, and the fans will slowly increase to 60%, making them less noticeable. In addition, this deals much better with rapid, spikey heat loads.
Mobo presets are generally shit. Modern CPUs actually run very warm at idle and low load, so a cuve that has too many points at lower temperatures will have the fans spinning up and down constantly. A better starting point for a custom curve would be something like >0C @ 20% RPM, >65C @ 40% RPM, >70C @ 50/60% RPM, >80C @ 75%RPM, >90C @ 100% RPM. For minor computer tasks the suggested curve will keeps the fans below 40%, espescially with 2-5 seeconds of ramp up. Fan quality/static pressure/etc. matters, espescially for noise-to-performance ratios, but the AIO radiator or tower cooler does the bulk of the work as long as there is some airflow. The thermal mass of a large tower cooler or AIO will handle temps spikes easily even with low fan speeds.
You can often simply set a fixed 30/40 or 50% fan speed that doesn't respond to temperatures at all and be just fine with a decent air flow case.
Learn to use a software temp monitor, run it in the background, and check the max temp. If you never hit TJmax for your CPU, you are golden and might even be able to lower RPMS. If you hit TJmax, you might want to increase the RPM, but first do a bit more research - resest the monitor right before gaming or other high CPU load task, then check during and immediately after - if the average temperature is too close to TJmax, you might need more/bigger cooling, less voltage, better case air flow, etc.
You can also use software to control fan speeds. There is at least one app I've used in the past that allowed for very complex curves and sources, including GPU. So I could have my bottom intakes that blow directly into my GPU only ramp up when the GPU got hot, while my AIO responded to CPU, etc.
*BIOS doesn't not read CPU temps directly, but rather from a senser in/near/below the socket. Software like HWiNFO64 will read it directly.
**It's near-impossible to damage a modern CPU via high temperatures as long as you don't adjust the protection settings in the BIOS. The CPU will simply downclock if it gets too hot.
***Mobos tend to push way too much voltage to CPUs by default. You can often use a negative voltage offset in the BIOS for vcore to reduce temps/power consumption and/or improve performance. Be sure to test for stability should you go this route.
****Mobos also tend to enable things like Multi-core Enhancement or PBO which may provide a very small performance increase in exchange for a large increase in power consumption and heat. I think PBO for AMD is nearly useless, except for the curve optimizer.