Where a home observation and an independent result disagree, the independent one is measuring something more specific. That does not make the home observation useless — it makes it a different measurement.
The rule of thumb is fine; the edge cases are where it earns its keep.