Nope, editing these values is not possible.
There was a recent change that allows datums to be excluded from the bounding box size calculations.
https://support.ptc.com/appserver/cs/view/solution.jsp?n=CS93540
It drives me nuts that I have to consider how I define certain datums since they influence how the model is zoomed in on. I currently have a model that I have to re-think because of some stupid arbitrarily rotated plane intersected with the mid-plane gives me the "horizontal" for that sketching plane. but this ends up being horrible because the intersection is 10x the longest dimension away from the default Csys. so when I put the model in perspective, or when I just zoom-all, the model is tiny! it's makin' me bonkers 
unfortunately no, it doesn't solve the problem if moving the bounds of the plane anywhere within it's planar space does not bring it closer to the model, still screwed.
though technically in some cases that could be a work-around.
so here's the basic case - 3 points constrain the curve. the curve is a spline that must be planar and NOT a combined projection for numerous reasons. The configuration below was how it started, a moderate angle, no problem.
However, adjustments to the design required a much steeper angle, which brought me to the following and this isn't even as bad as it was to be honest.
the final conclusion was to re-define how the sketch plane was constructed entirely. it isn't a problem, it's just tedious extra work for something that shouldn't be included in the bounding box if it's layered off.
bounding box for zoom-extents should be limited to only what is visible!
here was the final work-around which doesn't even require any hand-wavy nonsense of adjusting the plane's display-outline (which I'll admit can be very cool at times, but not very well parametric.
in the following 2 examples the horizontal planes are parallel. in the "new way" the plane is always stuck to the datum point but needs an extra datum plane to define it, But it is 100% parallel, and will never cause the bounding box issue.
new way highlighted
old way highlighted
for what it's worth, that display outline is relative to the internal origin of the plane. did you know there was one? yup, there is, and it's relative to how the plane is defined.
when defining a plane by 3 points, the order matters somewhat if you care how your planes are displayed.
I believe the origin of each plane is the first point picked. the second point picked lies on the x-axis, and the third point lies on the positive y side of the x axis. this matters for a number of reasons. automatic sketch orientation for starters. makes a difference how the tangent angle is projected for ISDX curves. oh, if you pick points A B C in counter clockwise direction the plane will face towards you. if you pick the points anti-clockwise the plane will face away from you.
and actually the tangent comes back to the original point which is - even in this example, all the parts of the planes that lay outside the boundary of the 3 points shouldn't necessarily be included in the bounding box. but especially if they are hidden.
