That is a good point. Personally I don't like puzzles or mechanics that are frustratingly difficult (or impossible) as they turn me away from the same types of puzzles in the future. YMMV though.
The 'problem' is that this particular playerbase has everything from innocent children to NASA like people so the puzzles are a mix of fairly easy right up to almost downright impossible....but someone solved it. Theres one involving Braille that blows my mind how they solved it. A lot can be done once you know the technique. Some I have solved and still have no idea what Im actually looking at or what to do with it or what it actually means so its over to Canonn website for the answer or the next step where I can pick it up again

But theres a fair few that are just follow the breadcrumbs type which anybody can do, even me.....then theres the Morse and the 'sound' ones and the pictograms, I cant even begin to keep up with those ones. Often they just need a key and once you have that you can do it yourself, or get so far anyway, sometimes the problem is knowing what format the key is in. eg the Halloween mystery, I knew it was Binary but had no idea how to use that info until someone posted the Binary key and then I could do it myself.
Some are logical and will be solved by trial and error. Carrying a Tharg object to open the doors is a logical thing to try and almost certainly would be attempted by someone for example. I love the fact that they also included the reaction from Guardian objects, the enemies of the Thargs, expecting someone to try that as well.
I think a few of them are designed or imagined to be solved by players working together, rather than one person working alone. Similar to the infinite monkeys infinite typewriters thing, eventually someone would try something that worked and mostly they have, the exception being in the next paragraph..
What staggers me is the confidence Frontier had when they wrote some of the puzzles that someone would solve them. All the UA stuff, the distance from Merope being the right 'number' to calculate by etc They had to give a hint eventually about 'listening to them' (UAs) to use Audacity or similar due to the story needing it but that solution suggests that they thought someone would stumble across it eventually if not even solve it within a certain timeframe. Makes me wonder what we have missed that didnt need a hint to move the story along and so they still havent told us or hinted at it.
Interestingly sometimes the simple ones take the longest as we overthink them. A fairly simple Rail code kept us occupied for 2 years despite being told all the info we needed was right there in front of us, people went searching elsewhere and overcomplicated it.
Personally speaking if I purely searched for the answer I would burn out fairly quickly so I dip in and out and keep it in the back of my mind in case i see anything that might be relevant and that way its a good distraction while still being fun.
And you post on the SC thread so that really is saying something
