DATA LIST FREE / X1 X2 (2F1.0) A1 (A501) A2 (A600).
BEGIN DATA
1 1 A B
1 1 C D
END DATA.


begin program.
import spss, spssaux
vardict = spssaux.VariableDict(variableType="string")
verylong = [v.VariableName for v in vardict if v.VariableType > 500]

longstr =  [v.VariableType for v in vardict if v.VariableType > 500]
#Grabbing string length

spss.StartProcedure("Over500")
table = spss.BasePivotTable("String Vars Over 500", "COMMON")
table.SetDefaultFormatSpec(spss.FormatSpec.Count)
table.SimplePivotTable(rowdim = "Variable",
     rowlabels=verylong,
     collabels=["Length"],
     cells=longstr)
spss.EndProcedure()

#if verylong:

#    spss.Submit("delete variables " + " ".join(verylong))
end program.

begin program.
import spss, spssaux

vardict = spssaux.VariableDict(variableType="string")
verylong = [v.VariableName for v in vardict if v.VariableType > 500]
if verylong:
    spss.Submit("delete variables " + " ".join(verylong))
end program.

Another sneaky way to do this as well (without Python) is to use ALTER TYPE  and SORT VARIABLES (although these were introduced as of V16 - so they won't  work going back really far).

What this syntax does is to change any variable of length 501 to 502 (which seems like a reasonable concession to me), then creates a variable to note the start of the string, this allows one to then sort the file and then drop all variables between the two variables you created.

**********************************.
ALTER TYPE ALL (A501 = A502).
STRING @Holder (A501).
SORT VARIABLES BY TYPE.
NUMERIC Temp.
DISPLAY DICTIONARY /VARIABLES = @Holder to Temp.
MATCH FILES FILE = * /DROP @Holder TO Temp.
**********************************.

If the ordering of the variables is important to keep you can use the VARIABLE ATTRIBUTE command - although that is not convenient like the Python solution.