When we first got our 6.5cm back in the summer I took some range pickup brass (believed to be 1x; no resizing marks on the cases when picked up) and made some lighter loads with IMR4895 (a close analog to H4895, but not exact) just to introudce the kids to it and to burn up what was left of some Nosler 123CCs from an old project. I used 38.0 grains.
Fast forward to last week and, piggybacking on what I'd learned with that load, I loaded up some 130ELDMs (also left over, from a different project) with 38.5 grains of IMR4895. I didn't have published data for either of those but put them together based on Gordon's Reloading Tool model predictions, which are generally accurate enough to be useful if you approach them carefully. Don't plug in some faster powder and go straight to a max load. Terrible idea guaranteed to eventually blow a primer when you least expect it. But I'd already worked with the IMR4895 with the 123 grain bullets and knew what to expect and also knew that the Tikka we're shooting has a long throat and that I'd be seating the ELDMs long, which reduces pressures.
I ended up with the 130eldm seated to the same length as my 147 load (~2.93") and they chronograph 2700' and hit to POI (or within less than 1/2moa of POI) of the 147 load at 225 yards which means they'll almost certainly be within 1 click of the 147 load out to 400 yards or further.
I shot a bobcat with this 130 load earlier this week. With scope turned up to 10x in very low light I was still very easily (with a 10# rifle) able to watch my impact on him from 75 yards and watch it fall over dead.
My regular 147ELDM load for this rifle shows a recoil figure of about 11.2ft/lbs. This load is right at 9.4ft/lbs, a very significant reduction, down near what you'd expect from .243 or 6CM type stuff.
If you wanted to go even lower, once upon a time IMR published data for a lot of typical centerfire cartridges using IMR4227 or IMR3031 or SR4759 or IMR4064 or IMR4320 and as a general rule, if you have any of those powders, the faster you go on the burn-rate chart, the less recoil it takes to produce a given velocity.
I have used IMR4227 to make some very light recoiling loads, just to prove that I could do it, I guess, with various calibers. You just have to be VERY careful with weighing your powder; when you're trying to stuff 45+ grains into a 6.5cm sized case an error of a single grain is difficult (because the case is near full) and insignificant in terms of safety. But when you're using maybe 30 grains of a much faster powder, a single additional grain could very easily cause a blown primer or worse.
Point being, you can do a ton of reduced loads with faster powders, but there's an extra level of care needed in load workup and assembly.
You could probably reduce your H4350 load by a grain or two. I'm saying that based on theory, not experience, and don't know how low you'd go before you started seeing erratic performance. Most powders seem to have some minimum pressure level where they work best in terms of SD and I don't know where that is for H4350 and don't know what to tell you there.
If I were you I'd look hard at IMR4895 or H4895 and also read through the reduced load data that Hodgdon publishes. But if you have any IMR4227 laying around (I don't even know that they make it anymore, my cans are likely as old as I am, picked up used but unopened long ago), it would be perfect, IMO.
TLDR: What other powders do you have on hand? If you don't like the price of H4895, what do you have on hand, or what can you get that's similar in speed to H4895 that's cheaper? And does $70 really even matter after paying for rotator cuff surgery?
