Indeed, you have uncovered a documentation error in the manual page you link to. However, you have also introduced an error of your own.
For your Mainframe installed SORT package (likely to be IBM's DFSORT or the competing SyncSort, but also possible the further rival CA-SORT):
//SYSIN DD whatever
Is where you put the control cards.
In the above, whatever can be *, DATA or parameters for an actual dataset.
The DDNAME for input to the SORT program you use is SORTIN:
//SORTIN DD whatever (same as above)
You accidentally made that SYSIN as well. It should be:
//SYSIN DD *
NEPTUNE
PLUTO
EARTH
VENUS
MERCURY
MARS
URANUS
SATURN
JUPITER
You should then see the output from the example in your SORTOUT spool file.
There are other DDs for when you do other things with SORT (like MERGE, JOINKEYS, OUTFIL) which can or do use differently-named DDs. It is also possible to override the standard names, but you would not be able to override them to SYSIN.
//SYSIN DD * is a bit like STDIN, but don't get carried away with the comparison. By convention, many Mainframe utilities use SYSIN for input. If a JCL-stream contains "cards" not preceded by a DDName, then a DDName of SYSIN will be automatically generated. COBOL has an ACCEPT verb for a type of input, and the default DD for this is SYSIN. However, simply including a SYSIN in the JCL for a step is no guarantee that it will be used. If the program on the EXEC does not use SYSIN, then simply including SYSIN is not going to cause data to be read from there.