Choosing the slower process on purpose
Picking the method that takes longer because of what it does not do.
Choosing the slower process on purpose makes sense whenever the faster method's speed comes bundled with a side effect the part cannot afford, because a process rarely gets to be fast for free, it is usually fast precisely because it delivers energy or force quickly and unevenly, and that same speed is often exactly what damages the very thing it is trying to make.
What is actually happening
Nearly every pair of competing manufacturing processes trades speed for gentleness in some form, a laser cutting quickly by concentrating intense heat that leaves a narrow zone of altered metal behind it, a fast mechanical punch shearing a hole in an instant at the cost of a rougher, more stressed edge than a slower drilled hole would leave. The faster process is not defective, it is simply doing its job by delivering a large amount of energy or force into a small area in a short time, and that concentration is exactly what produces the very side effect a gentler, slower process avoids by spreading the same work out, whether across more time, a wider area, or a lower intensity, until the material never has to absorb the same disruptive dose all at once.
This is not a rule unique to cutting, either, since the same trade appears throughout manufacturing wherever a fast method exists alongside a slower one aimed at the same result. A part cooled quickly after heat treatment can lock in stress a slower, controlled cooling schedule would have let relax naturally, and a fastener driven quickly under high torque can yield slightly at the thread in a way the same fastener, tightened slowly and checked as it goes, never would. Speed, in each case, comes down to a choice about how much the material is asked to absorb at once, rather than simply to time saved. The energy or force has to go somewhere, and a process that delivers it in a rush leaves the material no time to redistribute that load internally before the next dose arrives.
The frozen-meat comparison
Thawing a piece of frozen meat overnight in a refrigerator takes the better part of a day and demands nothing from anyone beyond patience, while thawing the same piece of meat in a microwave's defrost setting can finish in minutes. The microwave is not simply a faster refrigerator, it delivers its energy unevenly, heating the thinner, more exposed parts of the meat quickly enough that they begin to cook while the thick centre is still frozen solid, a side effect no amount of clever settings fully avoids because it is a direct consequence of how quickly the energy has to be delivered to finish the job fast. The refrigerator's slowness is the entire reason the meat thaws evenly all the way through with nothing at the edges ever crossing into cooked, a feature rather than a limitation to be tolerated, and anyone thawing a fine cut of meat for a meal that actually depends on its texture chooses the slow method deliberately, knowing exactly what the fast one would cost. A thick joint of meat makes the same point even more starkly than a thin steak does, since its greater bulk widens the gap between a fast-heated rim and a still-frozen core, meaning the very cuts that most reward careful cooking are also the ones a quick defrost damages worst.
Why the trade-off is rarely obvious from the finished part
The side effect of a fast process is often invisible on casual inspection, a heat-affected zone a few tenths of a millimetre wide, a slightly work-hardened edge, a residual stress pattern locked into the material, none of which show up as a visible defect on a part that otherwise looks completely finished. That invisibility is exactly what makes choosing the slower process on purpose look, from outside, like an unnecessary or even wasteful decision, since the faster alternative appears to produce an identical part in a fraction of the time. The difference only becomes visible later, in a fatigue crack that starts at a stressed edge a slower cut would never have left, or a hardened part that has quietly lost its temper near a cut a waterjet would have left completely untouched. Nothing about the part's outward geometry changes as a result, since the dimensions measured off the drawing come out identical either way, which is precisely why a purely visual or dimensional inspection can wave through a part carrying a hidden weakness that only a fatigue test, a hardness check, or years of actual service would ever have exposed.
The number that matters here
A part cut by a fast, heat-based process can carry residual stress and altered material properties within a band just a few tenths of a millimetre wide along every cut edge, a region too thin to see or measure without deliberate testing yet large enough, on a part with many cut edges or a part relying on consistent material properties throughout, to become the actual origin of a failure that shows up only much later and far from where anyone was looking for a cause.
What follows from this
Choosing a process on purpose, rather than by default, means asking not only how fast or cheap each option is but what each one's speed is actually costing the material to achieve it, and picking the slower option whenever that cost lands somewhere the finished part cannot afford to carry it. This is rarely a decision made once for an entire shop, since the same part might reasonably be cut fast in one revision and cut slow in the next, once a failure, a tighter tolerance, or a new understanding of how the part is actually loaded changes what the finished edge is allowed to hide. A part with generous margin and no fatigue loading to worry about can happily take the fast process's side effect without ever noticing it, which is why choosing slow on purpose is a judgement about one specific part's demands rather than a blanket shop rule.
My key error with this
The device needed a small set of gears, and printing them was the obvious move, since I could have a set the following morning and the loads involved were nothing dramatic. The teeth failed, repeatedly, in the way printed teeth always fail, which is along the layers that run across the direction the tooth is loaded in. I lost more time to reprints and to diagnosing an intermittent slip in the mechanism than I would have spent waiting on a slower process at the beginning. Having the gears cut instead, from sheet, where the material runs continuously through the tooth, took longer to arrive and then simply worked. What replaced the belief is that the speed of a process is only worth counting if the part it produces is one you can keep, and that when the fast option keeps failing in the same place, the honest reading is that the process and the part were never suited to each other rather than that the settings need another attempt.