For this I have no correct answer Melissa.
To digress, I believe Architeuthis to be quite closely related to the likes of Moroteuthis (another genus with 'large' species), Onychoteuthis and Ancistroteuthis on morphological, anatomical and genetic grounds. But there would be huge debate if I were to suggest it (so pretend I never wrote this, ok).
The paralarval forms of these genera, in different families (Architeuthidae and Onychoteuthidae) are almost inseparable through to mantle lengths of ~ 10mm. At mantle lengths smaller than this the two differ slightly in shape, and in chromatophore distribution.
The adult Architeuthis doesn't really differ from its paralarval form - it is like a giant paralarva, whereas at mantle lengths in excess of 12 mm Moroteuthis develops hooks on its tentacle clubs (the expanded distal-most part of the tentacle); in Onychoteuthis and Ancistroteuthis these develop earlier. Otherwise each has 4 rows of suckers on the tentacle club in the paralarval stage (and Architeuthis retains this character state through into the adult form). I think Architeuthis has a photophore (light organ) on the end of its ink sac, although it isn't supposed to have photophores (there's a very strange structure at the end of the ink sac in New Zealand specimens at least, that looks just like a photophore; should it prove to be one then this would render the Onychoteuthidae and Architeuthidae even more similar).
So, back to your original question, in a rather convoluted manner, speciation often happens in terms of size (instead of the acquisition of a new character or character state); as such you often find 'giant species' of this, that and the other - 'giant' being a relative term only. So the speciation leap from an ancestral Architeuthis to a modern-day Architeuthis need not have been through a slow increase in size over many many millenia, it might have been a very sudden speciation event (one day the species 'woke up' and was seriously big; its' next of kin could have been a comparatively small species).
The likes of Moroteuthis live to much greater depths than Architeuthis (to 1000m in New Zealand waters, whereas Architeuthis is generally found within the water column at depths of ~ 500m), so if the two are related or have relatively Recent ancestry (that's if I ever suggested that they were related .... and of course I would never do that because people would yell at me), then Architeuthis may have suddenly grown larger and concurrently moved shallower than its ancestor.
Of course there's another reason that Architeuthis might be big, and that's because it is quite a passive beast that drifts in the water 10-or-so metres above a large shoal of large fish (hoki), dipping its tentacles into the school of fish and then quietly withdrawing them, unnoticed, expending very little energy in the process.
The blue whale, another rather friendly beast, also grew large, eating a diet of comparatively minute animals, as did the whale shark (two more examples of gross gigantisism), so you don't really have to live deep to be big (and in fact the deeper you go there's probably a tendency for the animals to get smaller). Your pycnogonid (sea spider) example is a good one, because the largest of these brutes do live down deep.
Not sure if I've answered the question at all, but until Tony builds that time machine into the new Tonmo board we'll only ever be able to guess what really happened. There is of course an alternative to what I propose above, and that is that the ancestral Architeuthis was twice the size of the modern-day Architeuthis - the species could actually be shrinking in time (like my brain). Another alternative, a consequence of a shrinking brain, is that everything written above is nonsense.
Any debate is welcome. My belief is that Architeuthis is just a giant paralarva with serious hormonal problems, subject to premature maturity.
Cheers
O