Performance left on table by not skipping ahead to study start region

[alecloudenback]

Simpler codebase is nice. Expecting some performance regression due to

        while first(result).to < study_start
            popfirst!(result)
        end

In situations where from << to and we are doing some monthly study being able to avoid the extra work likely reduces computation time by a significant factor.

Originally posted by @MatthewCaseres in #24 (comment)