From the ISO 32000-1 specification:
Table 53, 8.4.1 describing initialization of graphic state at the start of each page:
The precision with which curves shall be rendered on the output device (see 10.6.2, "Flatness Tolerance"). The value of this parameter (positive number) gives the maximum error tolerance, measured in output device pixels; smaller numbers give smoother curves at the expense of more computation and memory use. Initial value: 1.0.
Table 57, 8.4.4 describing the "i" operator:
Set the flatness tolerance in the graphics state (see 10.6.2, "Flatness Tolerance"). flatness is a number in the range 0 to 100; a value of 0 shall specify the output device's default flatness tolerance.
Table 58, 8.4.5 describing the graphic state parameter dictionary entry
Number, (Optional; PDF 1.3) The flatness tolerance (see 10.6.2, "Flatness Tolerance").
10.6.2 Flatness Tolerance
The flatness tolerance controls the maximum permitted distance in device pixels between the mathematically correct path and an approximation constructed from straight line segments, as shown in Figure 54. Flatness may be specified as the operand of the i operator (see Table 57) or as the value of the FL entry in a graphics state parameter dictionary (see Table 58). It shall be a positive number.
It appears to me that the above clauses are referring to exactly the same thing. If that is correct, then the range and default value for flatness tolerance is ambiguous:
Either the default is 1.0 or it is 0: pick one.
Either the range is 0 to 100, or is a positive number (any value > 0): pick one.