I think that longitudes in the range [-360,360] should be accepted without complaint. In some applications even wider ranges for longitude make sense, e.g., when you're keeping track of how many times you've encircled the globe.

But [-360,360] overlaps the circle two times, which seems ambiguous. How is that different from [-4pi,4pi], which overlaps the circle four times? In that case I think it's better to not restrict longitude range at all, if we tend to think that a longitude outside [-180,180] is not a mistake. Thanks for pointing this out, I'll reason about it.

I moved all the warnings from Distance to Point, which allowed to simplify the validation code significantly. For the reference, the previous commit was 58a37a8.

I also decided to not complain about longitude normalization, because I believe there's nothing error-prone with that (unlike latitude).

I'll leave it open for a day to reason a bit about this all, but I guess it's ready as-is. I would be glad if someone could review (especially the warning messages, as I'm not so confident with English).

Ok, I refined one message a little. Now the warnings look like this:

>>> from geopy.distance import distance
>>> distance((170, 20), (170, 10))
[...]/geopy/geopy/point.py:85: UserWarning: Latitude has been normalized, because its absolute value was more than 90. This is probably not what was meant, because the normalized value is on a different pole. If you pass coordinates as positional args, please make sure that the order is (latitude, longitude) or (y, x) in Cartesian terms.  This will cause a ValueError exception in the future versions of geopy.
  UserWarning)
Distance(1096.3515545690054)
>>> distance(20, 170)
[...]/geopy/geopy/point.py:184: UserWarning: A single number has been passed to the Point constructor. This is probably a mistake, because constructing a Point with just a latitude seems senseless. If this is exactly what was meant, then pass the zero longitude explicitly to get rid of this warning.
  UserWarning)
Distance(3318.2210874060056)
>>> distance((20, float("nan")), (10, 170))
[...]/geopy/geopy/point.py:74: UserWarning: Point coordinates should be finite. NaN or inf has been passed as a coordinate.  This will cause a ValueError exception in the future versions of geopy.
  UserWarning)
Distance(nan)

Well, I hope it won't break anything. Merging and prepping the 1.13.0 release.