I may be the opposite to IchoTolot here, because I've definitely spent a lot more time creating levels than playing them.
Maybe this is the reason why my levels are easier than IchoTolot's on average, but given that the majority of New-Formats content seems to be up there with regards to difficulty and intermediate-level packs are still somewhat lacking, I don't necessarily consider this a bad thing.
Of course, if your goal is to "beat the player", as IchoTolot has stated in the past it is for him, then you want to get ideas that are as difficult as possible. I however see myself more like a dungeon master in a roleplaying game whose job it is to entertain and challenge the player, but not defeat them like in a direct competition between level designer and player. With that perspective, if you make too many levels too hard, you might just frustrate the average player in our already small community so much that they give up just halfway into the pack.
Designing the Groupie rank for Lemmings World Tour required me to play packs by a bunch of other content creators. This was like a bit of active research on my part; I deliberately played those packs to get inspirations for my own levels, because I wanted them to be similar to other people's levels and recognisable as such. But it's not like I didn't have enough level ideas of my own - otherwise, Lemmings World Tour wouldn't have had 280 other levels outside the Groupie rank.
I certainly agree with IchoTolot that, if you feel like you're currently out of ideas, playing other packs is a good starting point. I currently do this every now and then with Lemmings 2: The Tribes levels (both the original ones as well as Quest from Kieran 2). But sometimes it also just helps to take a couple of days off and allow your creativity to recover "naturally".
I definitely sometimes overdid it and may have created more level quantity than quality when I designed several levels a day.
Some users even said that one level per day might be too much. On the other hand, namida seems to be pretty quick at creating entire packs, so I'm pretty sure he had to make multiple levels per day, and those are of high quality and difficulty nevertheless.
When it comes to how I get my ideas, there are a couple of starting points:- recently, all my levels are song titles, so those can provide some framework, either through the title itself, or the story they tell, and/or the tileset(s) they suggest to be used
- I like to use somewhat obscure skill combinations for tricks, rather than plain-sight puzzles where all the skills are used in standard fashions and just add up to something that, for some reason, isn't as obvious to solve as you would expect it to be
- sometimes I also do "bottom-up" level design, starting with just casually and arbitrarily placing terrain on the screen, and then the ideas come only while I'm already building the level