I see a lot of statements about game design that treat it as technology, where there is objective improvement over time, or at the very least people talk in terms of "good design" versus "bad design". I think there are some things that can be pointed to as clear improvements and that you can talk about the quality of design. At the same time, I think mostly what people are describing when they invoke these things are trends rather than objectively better or worse approaches. I see game design a bit more like music than tech, mostly there are fashions or preferences that change over time, though there are also important changes in technique and approach that can be called improvements (the latter happens much less frequently than the former).
Rather than talk abstractly about this let me give an example of what what I mean. People will sometimes say things like "this mechanic allows for undramatic deaths in the game and therefore it is bad design", "this mechanic involves three steps rather than one, so it is bad design", or "this mechanic creates something outside genre expectations so it is bad". These are not good arguments. The first one could be bad design if the game was meant to avoid undramatic deaths, but the argument begins by assuming a certain set of preferences that not all games or gamers need to share. The second may or may not be a good argument, depending on how much you sacrifice for the simplification and what your goals happen to be. The third claim only matters if you are concerned about genre fidelity and just because something like Conan is a source of inspiration for a game that doesn't mean you want to follow all of Howard's conventions in terms of outcomes and internal physics.
Here is an an example outside of RPGs, in many modern boardgames it is simply taken for granted that systems where players can be eliminated from play for the remainder of the game, or where players can be crushed so badly early on they have virtually no hope of recovery, are instances of bad design because they are not "fun". I accept that this is a trend, that many people enjoy games that protect players from being removed from play and from boredom. But I do not accept that this is an objectively better approach than games that allow for such things. It is a trend. It is a fashion.
That isn't bad. There is nothing wrong with the trend. But it is definitely a trend and could change over time. Personally I don't like playing these sorts of games. I would take something like Axis and Allies, Risk or Samurai Sword over games like Puerto Rico any day of the week. I see why people like Puerto Rico and why they like modern Eurogame design (it certainly fills a demand). I just don't care for it myself and I know a lot of folks who feel similarly. I find the possibility of elimination increases my overall enthusiasm and excitement. True in a game where you fall behind and get bored watching others dominate the table, that can be painful when it happens to you...but knowing that is a possible consequence of play keeps me much more engaged and interested than if I am always assured a fighting chance.
Now some trends are permanent trends, some are passing. It may be that we don't see a game like risk take the hobby by storm for another two or three hundred years. More likely, people will become too familiar with the Eurogame trend and there will be another shift. It could be in a completely unexpected direction or it could look back at games that presently seem "unfun" but get viewed in a different light. This happens in all mediums.
Again, look at music. After the 80s, for years people shied away from twin guitar harmonies and rising staccato as grunge replaced what came before. The 1980s music had a baroque quality to it. There was complexity but also a strong formula. The 90s felt like a reaction against that, a desire to simplify and get back to something less predictable. We had the grunge era and it reverberated for some time, but then people once agains started drawing on the 80s. It would have been a mistake to discount twin guitar harmonies as bad music. Certainly I don't expect to see it used to the extent it was in the 80s any time soon, but it was a perfectly legitimate musical tool.
None of this is meant to discount approaches in game design that favor the current trends. It is useful to participate in trends because that is a good indication of what your audience expects and where it is going. The peril is in mistaking trends for permanent rules of what constitutes good design.
Presently it seems we are in a period where things have been moving more and more toward simplicity and streamlining. My own games tend to be simple and streamlined. I have enjoyed simple and streamlined games for ages now. But everything has a trade-off. And more and more, I find myself missing complexity, missing roundabout or multi-stepped solutions that give you more precise results. For all its strengths, simplicity does have its drawbacks.
So I think it is important to understand, that you are always in the middle of a trend. This can make it hard to see things clearly. And often it is the ones who see the trend, sense its faults, and break away from it, that create something truly new and interesting. Sometimes they do this by looking to fashions that have long been out of style, then refit them for a new context, sometimes they go in a wild and unexpected direction. Mistaking fashion and trends for progress can limit our potential as designers and players.