Intel has removed official support for S3 on mobile processors for multiple generations now. Would be stupid for a manufacturer to officially support S3, when Intel could very well rip out the support that is still technically there at any time. And that support seems to be only there, because Intel’s CPUs are still very similar across the line, from desktop CPUs, where S3 is still officially supported down to mobile CPUs like in the FW where it is not.
AMD seems to have removed official S3 support from their mobile CPUs as well, although I do not know if they still make a half-baked and bugged/untested S3 version available like Intel does. Since their CPUs are much more different between desktop and mobile, they’d have more reason to not support S3 at all on mobile CPUs.
Then, Microsoft says, switching between modern standby and S3 is only supported with a reinstall. Even back when Intel still officially supported both and Thinkpads had a BIOS switch to choose, which of the 2 the BIOS communicated to the OS. So any and all bugs you find with S3 on Windows might also stem from the fact, that you used that registry hack and cannot do a clean Windows installation for S3. I have not found any information, on what specific Windows features would depend on this and are only configured during the initial installation though.
You can think of modern standby what you like. I am aware of / have seen issues on FW like the USB4 and audio drivers (or others on other notebooks) bugging out and not sleeping correctly, which increases power drain in sleep significantly and is only fixed by a reboot. But nothing that will make the laptop get hot or drain its battery completely. A working, default Windows installation should a) doze to hibernate after it consumed 5% battery capacity sleeping and b) even if Windows does crash (which happens far too often, might be a FW firmware issue or a Windows issue, don’t know), it should shutdown instead of remaining on forever. I.e. even when a driver misbehaves and wastes power, it should still at most cost you 5% battery. The only question is how long the device will wake quickly from sleep, before it switches to hibernate, where waking will no longer be instantaneous.
Another issue that has plagued many modern standby devices is, Windows somehow missing that the device was unplugged from power and behaving like on AC power. Which means doing all kinds of background work. That is what usually caused the battery to be completely empty and the device hot. Because Windows thought mistakenly, that it had unlimited power. My old laptop is affected by that, if you sleep it and then unplug it from power. But I cannot reproduce this on my FW 12th, so that issue seems fixed. In fact, my guess is, that part of the reason why the FW wakes up by itself, even from S3, is, so that Windows gets notified of battery changes so it does under no circumstances miss that it was unplugged from AC power. Or similar, other devices that were designed and configured for modern standby, where a wake-up event is sensible to notify the OS of a change it might want to react to, but can also be handled and ignored in milliseconds. The same is basically impossible with S3.
Other than that, the FW seems to have an inefficient, but passable modern standby implementation. The whole reason for modern standby is giving more control to the OS and the normal drivers the OS uses instead of firmware that simply can shut power off or leave it on and cannot be reliably controlled by the OS. So most of the issues are OS & driver issues. And OS issues are more or less shared across all devices. Driver issues probably as well, at least for the more popular devices.
The biggest cause for varying hardware/firmware-caused power drain with the FW 12th seem to be the non-USB-C expansion cards, at least on released 12th gen firmware (I have been running the beta BIOS since its launch). The base power consumption in sleep just seems to be higher than with competing devices (seems to be down to the board design. Things like user-upgradable RAM might even play into that, hard to know how justified the power draw is compared to other notebooks).
Edit for completeness:
Since the notebook in modern standby is not technically off, the fan will keep running if it was running before you put it to sleep. But since it is no longer doing much, it should cool down and finally shut the fan off quickly. Windows’ is supposed to limit CPU power consumption in a way, so that the fans do never come on again. And this has been true for all devices I have had access to, except in the 2 cases I mentioned. Should Windows for some reason bluescreen and reboot, the fans might come during the reboot until Windows is in control again or gives up and simply shuts down. Or when Windows thinks it is running on AC power and does not do any of the power limiting. The 5 % battery limit Windows gives itself by default has been working perfectly for me on every device, if it is active.