Flex is a product of thickness, not hardness. A piece of steel will flex the same amount at Rc 50 as it will at Rc 65. The difference is that when the elastic limit is reached, one will take a set (bend permanently), and the other will break.
Fillet knives generally need to be thin. I use .065" stock for most of mine. They flex a lot. I use .100" stock to make a "stiffer" knife. I haven't used .120" stock for a fillet knife in many years....since I got tired of grinding it down to half its thickness. Mainly I use CPM-S35VN for these knives and HT to Rc 61-62. The get wicked sharp, the edge last well, and it isn't insanely difficult to resharpen.
Fillet knives are not intended to cut hard things, just flesh and soft bones. They are generally made around Rc 62. , The bevel should be a FFG ( full flat grind) to a near zero edge, with a fairly acute secondary added to that. I do the secondary around 10° . The amount of distal taper will control how much more the tip flexes than the blade's main body. A very thin tip will curve around ribs in filleting flounder, and a stiffer blade can be used to peel skin off a rockfish.
Since these blades are thin to start with, the average is about a 25% distal taper. If you are working from thicker stock, I would suggest 50%. BTW, I measure the distal taper 1" back from the tip.
I used to grind the blades and do the HT. Now I profile them, drill any holes, and send them to Peters in batches. I do the bevels after HT. I quit grinding at 220 most times, but on smaller blades, go to 400 grit. I sharpen the edge on a 400 grit belt. The satin grind looks good, and seems to be fitting for a working blade.
ADDED:
"Spring Temper" is a lower hardness, not a higher one. In making springs, they need to be able to flex farther without breaking. Thus they are tempered at a higher temper than a blade will be.