While I haven't had the opportunity to get in any 40k in quiet a while I've been able to get plenty of Magic in over the past month and a half. I've been getting my feet wet in Modern and learning about archetypes, the extent of netdecking and how to play better through being punished for misplays. Chatting with Godfrey, Son of Horus and Corvus about getting armies built for when or if I find people to play against has gotten me thinking. There seems to be an opposite mindset between the two gaming communities regarding netdecking or listing as the case may be.
One of the fundamental differences between the two games is that MtG has different formats, each of which has its own set of legal cards that players may choose from while 40k has codices which each limit the number and types of units a player can choose to use. Each game has a strictly limited resource to be allocated, points and cards respectively. However, these are very different in that any card equally consumes a slot in the maximum while models consume varying amounts of points. This leads to very different set of choices where cutting a traveler's amulet leaves you room for a cryptic command, but cutting a thunderhammer terminator frees up room for five fire warriors. In this, 40k allows for a much clearer metric by which to evaluate the choice you are making.
To try to make a clearer argument, let's compare codices to deck archetypes in a given format. In competitive 40k and MtG most lists out of a given codex or archetype look largely similar. The reason for this is simple: those combinations of units or cards win in the large sample you see in large competitions. The good thing about these large competitions is that in a large sample you get lots of people and each of them brings their own idea of what is good to either win or be eliminated. This gives us a better idea of what works than a smaller local tournament would. We can get a better idea of what is a strong army list from Feast of Blades then we could from our tournaments at the Keep simply because at FoB the winner had to play against more players with more diverse armies.
Going back to the netdecks, we need to ask why people decide to copy verbatim the decklists that win major tournaments. On the surface it's a simple answer: they want to win so they copy what wins. On a deeper level, the community and secondary industry supporting MtG has grown to such a size that you can get watch major tournaments live with color commentary and see not just what the top players are playing, but what decisions they make and how they make those decisions. This amount of information allows players to have a solid understanding of what their new deck does well and poorly. It allows for a starting point for people who don't fully understand the metagame.
Let's look at the equivalent behavior in 40k and then put it back in perspective. Lists like Leaf Blower, Flying Circus daemons, wave serpent eldar, etc. are generally highly repetitive in what units they run. This means they have fewer thought processes to evaluate as one daemon prince will function identically to another but will function differently from a greater daemon or a squad of daemonettes. A player who is new to an army will be able to grasp more of their army's interactions more quickly if they are identical. This is good for a new player, but can only develop thought processes so far.
Pick up that question I asked you to hold onto earlier. Look at it again. Unless you're quite well off or have very few opportunities to play, the big events where netdecks and lists originate from don't make up the majority of your gaming opportunities. For most people, their gaming is mainly going to their FLGS or a friends house and playing against their friends. Most of us have a small pool of regulars that we play against and thus we have a smaller sample size. Where there may only be one rogue deck or list that wrecks that 'optimal' list that won that event in every hundred or thousand players in the nation, it doesn't change the fact that you may be playing against three friends who play it. If you play that optimal list in your play group don't expect many wins.
However if you're playing that optimal list because you're new to the game or army and decide you want to win more than you want to play that list, you can and should change it. Where a wave serpent army might have a horrible time dealing with a nob biker or thunderwolf cavalry list if they only run fire dragons and dire avengers, they might have a much better time if they brought in some wraithguard even if, by whatever measure, wraithguard are worse than the unit they're replacing at a large tournament setting. The immediate gain is that the wave serpent player manages to kill those annoying bikers, but the long run gain is that they learn how another part of their army interacts with the rest and becomes a better player.
In summary, netdecks exist because we see what works at large events, we want to get the same results and thus we use the same list. That may work at large events, but we mostly play in smaller events and playgroups. This means we have a smaller sample size and can have a much more varied result. If we stay to the list that won the large event we stunt our growth and chances for winning relative to if we would adapt to our local playgroup. Additionally, understanding how and why our 'optimal' netdecks don't work in our playgroups will make us a better player and allow us to personalize our armies for fun and profit.