This.. sounds interesting.
I personally go for 100% always. In the long run maybe you come out ahead with a >50% strategy if you incorporate it at low skill level gains but when you're using 10+ SR weapons for a skill up and fail, you are costing yourself a big chunk of time compared to just using the 14 initially.
Subjective I guess. I just look at it as a fixed cost for skilling up weapons. I haven't lost a single one to a failure.
But each to their own!
As long as you don't go OVER 100%, on average you'll end up the same regardless of what chance you decide to go with. 50% is going to be the most variable, so it's possible to get more swings in your favor, but it's also just as possible to get as many against.
The only thing you shouldn't do (statistically) is go above 100%.* If you're stuck at 95% or whatever and everything will push you past 100%, just go for the 95%. Yes, failing that means those weapons are wasted, but even if you succeed at >100% you've wasted weapon points regardless.
Of course, if you're just using an extra R weapon it's not a big deal if you really want that insurance. I'm just saying that on average you'll end up using more fodder overall.
* The exception to this is when using skilled up SSRs, since they essentially give you "free" points at skill level 2 or higher, so it's okay to go past 100% with those.