Less is more, and spending the same time creating a lower amount of difficulties with high quality than creating a huge amount with lower quality is worth a lot. Getting a lot of GDs is likely to reduce the quality of a mapset as a whole, since GDs can make a spread less polished (lower difficulties) or lead to very similar maps in one mapset (higher difficulties), which is worthless for the players (not considering getting double PP) and just creates more work for the modderns and Nominators checking the mapset.
I personally follow this mindset for my own mapsets and I think it's reasonable, but I don't think the 8-difficulty-limit will necessarily lead people into this mindset. Instead, it might just demotivate them as the general consensus from the community seems to be. This is why I personally suggest making the hard limit a guideline at best.
Instead of limiting the amount of difficulties people are allowed to make, we should try to think of ways that motivate people to map fewer difficulties, apart from the fact that it will be easier to rank the mapset. A possibility would be to allow for higher difficulty gaps between two consecutive difficulties (preferably for longer songs, which would solve some other mentioned problems, too) as long as the spread is linear, to encourage people to make small spreads that still cover a wide range of skill levels.
I personally follow this mindset for my own mapsets and I think it's reasonable, but I don't think the 8-difficulty-limit will necessarily lead people into this mindset. Instead, it might just demotivate them as the general consensus from the community seems to be. This is why I personally suggest making the hard limit a guideline at best.
Instead of limiting the amount of difficulties people are allowed to make, we should try to think of ways that motivate people to map fewer difficulties, apart from the fact that it will be easier to rank the mapset. A possibility would be to allow for higher difficulty gaps between two consecutive difficulties (preferably for longer songs, which would solve some other mentioned problems, too) as long as the spread is linear, to encourage people to make small spreads that still cover a wide range of skill levels.