In theory, some edges are small and prevent you from taking larger edges later. It would be ok to pass up those edges. However, in practice, it is very dangerous to pass up edges you have identified.
One reason is that the edges may not be small. As poker players, we usually focus on what the right play is, not by how much. Poker players tend to be bad at determining which hands are profitable by a lot, and which are profitable by a little. The result is that people use the idea of passing up small edges to justify passing any hand with which they feel uncomfortable, and they give up opportunities which are very profitable. They pass up situations which are more profitable than playing normal winning poker.
Another reason is that many edges do not prevent you from taking later edges. In fact, taking edges now may increase the edges you can take later. Imagine that you know that you will get to double everything at no risk in 1 minute. How much should you wager now as a 60:40 favorite? To maximize your profit, bet as much as you can now, so that you will have more money on average to double. If you try to gather chips early, then sometimes you bust out, but when you succeed you allow yourself to take advantage of many more good opportunities later. Particularly when you raise rather than call, and when you play a small pot rather than a large one, taking a small edge now can allow you to find more edges later.
It is common for people to say that they want to have a low-variance style. They want the wins without the frustrating downswings. In NL cash games, that might be possible, since your playing style can have a large effect on your standard deviation. However, in SNGs, huge differences in playing style only have a small effect on your standard deviation per tournament. If you decide not to call all-in for 2000 chips on a borderline decision, you don't finish the tournament with 2000 chips. You will still have to gamble later. Not calling primarily delays the gambles rather than avoiding them. Players who try to adopt a low variance style in SNGs often hurt their ROIs significantly while reducing their standard deviation per tournament from something like 1.54 buy-ins to 1.52 buy-ins per tournament. The result is an increase in downswings, the opposite of the stated goal. At the table, focus on winning, not on futile attempts to eliminate variance.
In the past, the games were softer. For example, people argued about whether it was possible to sustain an ROI of 50%. Now, many people hope that they have an ROI of 5-10% even in low stakes games. If you find a game where your ROI is over 20%, then maybe you can pass up some small edges. However, if your ROI is under 20%, your main focus should be identifying edges and taking them, not passing them up. The players with the best ROIs get them by recognizing good situations better than other players, not by passing up edges.
