Larger Spectral Norm Benchmark Case

[insertinterestingnamehere]

This adds a larger spectral norm benchmark case since the existing problem sizes result in near trivial running times.

[insertinterestingnamehere]

One potentially open question on this one: is the "submitted" directory the right place to put this? It's running on a submitted code, but I wasn't entirely sure.

[jabraham17]

If you don't want to edit the submitted directory, I believe test/studies/shootout/spectral-norm/bradc/spectralnorm-blc.chpl is an identical perf test that is not in "submitted". Any "submitted" test likely has a copy somewhere else in our test tree

[insertinterestingnamehere]

FWIW, I don't mind editing the submitted directory. I just wasn't sure what made sense in terms of the current directory structure and wanted to confirm. Really whatever is fine with me.

[jabraham17]

I'd probably edit the other benchmark, not the submitted one.

Also, is the intention of this PR to add a perf graph for the larger data size, or just run it with a larger size?

[insertinterestingnamehere]

If I understood right, I think Brad wanted to add another graph to maintain continuity for the old one. I just want at least one that takes a less trivial amount of time. All of the spectral norm benchmarks run on relatively small problem sizes right now.

[insertinterestingnamehere]

If it were up to me, I'd change all of them to just run on the bigger problem size and drop the size loop that's present, but I'll definitely defer to you all on what's wanted there.

[insertinterestingnamehere]

Alright, just switched where the new benchmark case was added.

[insertinterestingnamehere]

Although, looking at this again, I think this size may be slightly too big for smaller hardware. I'll scale it down a little bit.