I start by picking a bullet that should do the job required, whether its punching longrange paper, vaporizeing rodents, killing coyotes, club buffalo or whatever. Seldom is the sole purpose of the rifle to try to put bullets close together at 100.
From there I'll pick a powder with a reputation for getting the accuracy and speed that I want. Starting powders might be picked from Nosler, Barnes, Cooper or my buddy Rob's favorites. With your .204 I would start with Benchmark, just like I did with my own.
If the bullet can be seated to the lands lightly while still working in the mag, I'll start there and load the whole sequence from minimum to maximum in whatever decent brass is available for that caliber. I used Nosler in my .204. Somewhere in that sequence it will either shoot or I'll scrap that bullet. If I find a wide band in charge weights where it groups well and shoots the different charges to the same POI, I might play with the seating depth a little. If its picky, the load gets scrapped. A more organized person would be doing all this over a chronograph, but I'm too lazy. If the load proves up I'll clock it now, do a preliminary drop chart and take it out to the longest range I expect to use it at.
I'm finding that it is easier and cheaper to plan for sucess from the beginning than to hope to find it through some random collection of components. These days that often means that the process started by picking the barrel. Stick good bullets in good barrels and good things tend to happen. A rndom search can quickly use the price of a barrel in wasted components, and with some cartridges it can also use up a significant amount of the barrel life.
If I find a useful load for a purposeful bullet I'll try to resist the urge to find out what else the rifle will shoot.