Why do we reap? One of the biggest benefits of reaping is that reaped nodes are not searched by Super Search.⁽¹⁾ They also do not show up on the lists of postings in the various sections of the site, e.g. Seekers of Perl Wisdom and The Monastery Gates. So unless you really try to drill down to a reaped node's original content, you will never see it. (Reaped nodes are included in Newest Nodes, if you've enabled that in your User Settings.) So, other than the content being kept "near line" to slake the appetites of the morbid, reaped nodes are as good as gone.

Nuking a node removes its record from the perlmonks database, so it is really gone, as far as the users are concerned. However, the data is not lost entirely — it is copied (serialized) to a backup file. So if the gods should deem it necessary, one of them can restore the nuked record. But note that this process is substantially manual, so requests from users to "please un-nuke my node" will likely fall on deaf ears. (And if I'm mistaken about this, I'm sure the gods will strike me down.)

There has been a movement/desire amongst the janitors for quite some time, to have janitor-nuke changed to janitor-reap. I'm unaware of any dissent on the question. So I'm a little suprised that it hasn't happened yet.

But of the question, "Why do we need janitor-nuke/reap at all?" It basically boils down to the fact that the standard reaping mechanism, which is automated (and based on the whims of hoi polloi), sometimes doesn't do the right thing, in strange/corner cases that really require a human's good judgement (where the human is someone who has the PerlMonks tenure, experience, and wisdom to know what's right for the monastery).

(1) Technically, the reaped node's content is replaced with standard text, which, besides a pointer to the node's original content, includes the text of the consideration reason which led to the reapage in the first place; and this content is searched by Super Search.