Mathhammer is a tool often used to evaluate in game decisions and unit choices. Although crunching numbers is relatively easy in-game, where you know the exact situation and variables, mathhammering during your listbuilding process is often viewed as equally productive. However, simple choices such as the positioning of your models can completely change around the expected outcome (this is especially true in close combats, where having to charge through cover or get singled out into a challenge can make all the difference). I'm going to spill out some thoughts in no relevant order on this very topic.
Mathhammer and listbuilding most often coincide with netlists. These sorts of lists have been built using the theoretically most powerful and cost efficient units in a given codex, maxed out. This will give the list builder the impression that his list will have the highest probability of success. In my experience, these lists often rely on certain factors going their way. The mathhammer has been calculated expecting certain variables. Not only this, but the lack of diversity means that there is a lack of diversity in the threats you'llbe facing. A more uncoventional list will most likely include choices that are not the most cost efficient choice on paper, inclining that the listbuilder has not been solely (or even primarily) relying on mathhammer. Netlists are often more one-dimensional.
What does this mean? As an opponent, you have a clear idea of how that netlist is going to try to win. I've often read on the internet of evaluating your opponent's most likely path to victory and countering it (I think I first read this on 3++). With a spammed out list, this is easier. There are few netlists that are truly flexible. The units it spams may be multipurpose and effective in more than one phase of the game, but they are still the same unit, with the same characteristics.
Once we begin to consider how we are going to counter our opponent's path to victory, it is equally important that your own list has more than one way to play the game. 40k has at times been called a game of rock paper scissors as different spam lists duke it out and hope to come out on top in the round pairings. This is because the less different choices you have, the less variables you control as you are restricted to a certain type of play. Differing deployment options, range bands and mobility will dilute your list in a certain field, but allow you more avenues to take over the course of the game. This is especially true now that 40k is primarily a shooting game. Take away 2 turns of your opponent's shooting and the expected amount of damage he will deal over the course of the game reduces drastically. With more units left towards the end of the game, you may be left with more options to contest, control or saturate certain parts of the battlefield.
Although I feel that theoretically a less spammed out list is better than copy paste, we have to remember there are way more factors when it comes to list building. Some playstyles lend in to aggressive dominance of a certain phase, it's really more about what kind of a player you are. However, it is important not to be blinded by mathhammer both when building your own, and evaluating your opponents' lists. There are many variables that can turn the tables, synergies that may not reveal themselves at first glance.