So let me start off by saying that I recognize that there was initially a genuine problem with people who didn't want NSFW content being exposed to it.
Some of this was due to the fact that not all content was being correctly flagged as NSFW, and some of it was because a lot of users didn't realize that individual users can choose to completely block an entire instance - which is not only a very easy and fast solution, but also does not require an all-or-nothing approach of defederating from NSFW instances.
A number of changes were made, but some of those lingering changes have meant that people who do want to see NSFW content are not because:
-
Even having subscribed to several NSFW subs, they are effectively completely missing from my feed.
-
Most NSFW thumbnails are blurred.
Both of these behaviors should not be occurring if a user has chosen in their settings to NOT hide NSFW content.
However, I will also say that the blurred state is something that deserves its own user setting (i.e. so that a user can choose to NOT hide NSFW, but still want them blurred or not) - preferably with the granularity to set it for various sub-types of NSFW (e.g. porn, gore, etc...).
The downside is investing tons of time making and testing code that ultimately accomplishes nothing that wouldn't also be achieved by making your own instance.
Except here it's not about writing a new functionality as much as getting the UI to do what it says it does.