Krabbe draws on the framework set out by van Eemeren and Grootendorst. He refers to work of theirs that is earlier than the paper you have read (so what he says on p. 253 about their classification of ad ignorantiam won't fit what you've read); but the numbers of the rules he refers to (in Roman numerals) matches the numbering of the list on pp. 135f.

Krabbe's account looks at ad ignorantiam in three ways in the three sections on pp. 254-261. You should think what the three are and how he sees them as related. In describing the first, he locates what he calls a "dialectical shift." As you work out what he means by this, it may help to think why he says the fallacy in the case he describes could also be characterized as a straw man fallacy. (If the term "straw man" is unfamiliar, look in the index and, in particular, at the fallacies van Eemeren and Grootendorst associate with their rule 3 on p. 139.)

You may ignore the more detailed discussion of the third form of the fallacy on pp. 261-264. If you do look at it, it may help to know that Krabbe uses the notation ":=" in assigning specific meaning to symbolic letters. His English readings "becomes" and "becoming" may only disguise the fact that this notation is frequently used in progamming languages to assign values to variables. So "x := 2" might be read "let x have the value 2" and Krabbe's uses of the notation can be read in an analogous way. In cases where "E" or "R" appears among other symbols on the left, that is the symbol that is being assigned a specific meaning.

Krabbe concludes by returning to the example from Spinoza that he presented on pp. 253f. Whether or not you find his account of it helpful, it is worth thinking for yourself about this argument and in what sense, if any, it is an example of ad ignorantiam.