I initially like the "compared to the same day last month" feature but I'm finding it to inaccurate to use for two reasons:
1) It doesn't take into account that last month started on a weekday but this month started on a weekend, so this month looks worse by comparison. It would be more useful to have a comparison to the last 28 days in that case.
2) If today just started it will still compare to the full day last month. This is especially noticeable at the beginning of the month when an entire days worth of revenue can have drastic effects on the percentage.