The occasional (but expected) outlier forces the scale on these graphs to be so large that normal figures (the other 99%) that have much smaller variance between themselves are difficult to discern without mousing over them to get the actual figure from the tool tip. It is not so much an issue of trying to know what each individual time point's value is, but to be able to detect a trend. The lines start to look rather flat, when a large outlier is present.
An example is Net Change in Cash Flow when a large purchase of aircraft have been made in the preceding 40 weeks.
This problem is typically handled elsewhere by presenting the data on a logarithmic scale. Then, the values closer into the median will "pop out".
Would be nice for the system to one of: 1) be smart enough to know when it needs to present a log scale, or... 2) have the option to default and/or toggle between log scale or not, or... 3) default to log scale all the time.
Since the graphs are provided by a third party package (from what I understand in response on another forum post), I imagine that they likely have the option to display this.
Not a huge issue, but a nice to have enhancement if that part of the code is being worked on in future.