Years ago, we ran titanium target tiles. And the problem was, the sizes of the tiles varied, both between part numbers, and even the (customer supplied) material for a given part number.
For example, most parts finished at 1" thick, but the stock would come in at anywhere between 1.08 and 1.35" thick. We were feedmilling these on 2 different machines, one on a 50 taper horizontal with a 5" iscar feedmill (120RPM at 60ipm! i think 10 flutes, .040 deep), and a 3" feedmill on a box way 40 taper.
One day They might give us 10 plates, 2 or 3 different sizes, various thicknesses. And we just set it so we called up the face macro after probing. Set size, tool diameter, xy dimensions. The probe would probe several areas in Z and find the high point, and do the math from there.
Another macro program was set to mill around the plates. Plates either had a straight wall, or 1 or 2 steps of different sizes. The macro even deburred.
It was easy, made quick work of it all. And we didn't have to reprogram anything. I'd guess we probably did about 100 unique parts like that.
So I guess I look at people saying that it is STUPID to have a macro for facemilling and can't figure out why. If you gave me 20 part numbers, I could generate facemill code for each one very quickly. But if I want to train someone to run parts at the machine, and all they do is facemill, then teaching them to change 5 numbers in a macro program simply makes sense. Don't have to worry about them knowing how to program, etc. Change the numbers, done.
What gives, guys?
Now I didn't review the code very fast, but one thing stood out to me. You say that the feedmill doesn't appreciate conventional milling? Unless you're doing it on a tinkertoy linear way machine, the feedmill shouldn't really care. You can bury the thing full slot and it shouldn't be much different than doing an 80% climb or 80% conventional cut. What machine are you running this on?
Edit: I'm by no means a feedmill expert, but I think I've read that feedmills which have straighter edges, rather than a sweeping radius, are better for titanium. There's "less" engagement of the insert in the cut, if that makes sense? At least that's what I think I read.
We use Iscar feedmills. These ones:
https://www.iscar.com/eCatalog/Family.aspx?fnum=2591&mapp=ML&app=65&GFSTYP=M