Further inconsistencies in Pickles dataset

[teutoburg]

• Temperatures listed in the Pickles table are not those listed in the original Pickles paper (Table 2).
• There is a M10III star in the Pickles dataset, which is truncated to M1 in the table. Its temperature and mass are also mismatched with that of an M1 star, however the filename correctly shows "ukm10iii.dat". I suggest removing this entry, because the spectral type designation M10 is dubious anyway, and M9 is close enough for all our practical purposes.
• Some of the spectral types in the original Pickles paper (Table 2) represent ranges of types, i.e. averages, e.g. B5–7 V or A4–7 IV. Those may IMO be written as the average of that range (i.e. B6V for B5–7 V or A5.5IV for A4–7 IV), as that should work just fine with the interpolations that are used to find the closest matching type in the data set for any given spectral type. However, currently those cases are again truncated to the start of that range, which is a bad representation of what they actually encode.