The proposed interpretation of nonpositive years is not arbitrary, it is a natural extension of the same formulas that we use for positive years.  On the other hand 1-9999 limits are arbitrary.  If we wanted to restrict calendars to 4-digit years we would limit years to the 1000-9999 range.

Note that interpreting year 0 as 1 BCE is not without a precedent.  ISO 8601 standard does exactly that with the caveat that "values in the range [0000] through [1582] shall only be used by mutual agreement of the partners in information interchange." Furthermore, "an expanded year representation [±YYYYY] must have an agreed-upon number of extra year digits beyond the four-digit minimum, and it must be prefixed with a + or − sign instead of the more common AD/BC (or BCE/CE) notation; by convention 1 BC is labelled +0000, 2 BC is labeled -0001, and so on."  See <https://en.wikipedia.org/wiki/ISO_8601#Years> citing ISO 8601:2004 sections 3.4.2, 4.1.2.4 and Annex B.1.1.

Furthermore, documenting the existing limits is not an easy task.  For example

>>> from calendar import *
>>> cal = TextCalendar(1)
>>> cal.prmonth(1, 1)
     January 1
Tu We Th Fr Sa Su Mo
                   1
 2  3  4  5  6  7  8
 9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31

displays 6 days in December of 1 BCE as blanks.  Similarly itermonthdays() and itermonthdays2() generators yield objects corresponding to these days.  Are these days within the calendar limits or not?

Note that I am not proposing to extend the range of the datetime.date objects.  Doing that would be a much more difficult task, but defining calendars for out of range years as those for year % 400 is trivial and much easier than to figure out and document the edge behaviors.