A few things to chew on for those who like to think along:
Wave drift forces, are forces due to absorption or reflection of waves. Waves carry momentum, and stopping or reversing that, will cause a reaction force. Structures with a small waterplane area, like spars or semi-subs, do not reflect a significant amount of waves; they leave the overall wavefront more or less undisturbed. This is however not the case for large ‘floating island’ type structures, or breakwaters. The wave-drift forces on an (approximately) infinite wall can be calculated from the formula found here. So substituting some numbers, for a breakwater or island, placed in clubstead ocean conditions with a significant waveheight of up to 8 meters (4m aplitude), this translates into almost ten tons of force per meter of structure exposed to the waves.
In shallow waters, these type of forces might be combatted successfully by means of plentiful mooring, although even that is stretching todays technology (link). In deep waters, mooring against these forces does not seem plausible at all (link).
A single ton of thrust requires about 60kW of engine power, which means designing for absolute positioning against these forces by means of powered thrust is not very plausible. Even if tugboats were working at full engine power neck to neck along the entire length of the structure, they still wouldnt be capable of even holding the structure in place.
So absolute positioning does not seem likely, but what about lazy, time averaged station keeping? The dependance of wave-drift forces on wave height is quadratic, so under more typical calm weather conditions, these forces would not be nearly as bad, nor would their time-averaged effect be. Yet given that these large structures would act as a wavebreak, there will be no significant waves on the leeward side. The island will move with unstoppable force, but the potential neighboring seastead island it is moving towards, will experience no such impetus. Hence, having mutiple such large islands drifting in relative proximity would be complicated, to say the least.
Your thoughts? Is this a dealbreaker for any wavebreaking type of structure? Or could these problems be dealt with somehow?