Current top of the line lithium ion batteries achieve around 250 W/h per kg, top of the line ships like the tripple E class with a carrying capacity of 165.000t would translate to ((250 x 1000 x 165.000)/1000.000) 41.250 MW/h. Let’s be generous and say that the advances lead to 100.000 MW/h, which is most likely decades into the future, the sun power cable is supposed to provide 15% of Singapores electricity, their current production capacity is 12.000MW, so 1800MW. This means you would need about one ship at the current capacity and about 0.5 of the hypothetical one per day, just to cover 15% (!) of their electricity demand and those ships would require power too, block valuable port docks and be prone to accidents, which would be a huge issue as this is critical infrastructure. According to google it takes about a week for a container ship from northern Australia to Singapore, so you would need to maintain about 14 or hypothetical 7 (plus some for reserve/emergencies) ships to replace this one cable. According to the internet and some guy on Quora a kilometer of deep sea cable costs around 1 million $, their are 4200 km of cable according to the projects website, which checks out with the 20 billion $ price tag of the entire complex including the solar and battery storage. Which leads me to estimate that the cable will cost approximately 4.2 billion $, however the operational cost will be very low and its reliability much higher. Submarine power cables have an designed life expectancy of 25 years, so your ships need to have an levelized cost of less than 12 million $ (14 ships)/ 24 million $ (7 ships) per year. New 20.000 TEU container ships seem to cost around 200 million $ at the moment and have an life expectancy of about 10.5 years, so the build cost alone will already blow the budget even if you sell all the stell scraps afterwards, this also excludes the cost of the additional batteries required. So I would say it’s safe to assume that battery ships will not replace sub sea power lines connecting mayor economic areas, maybe it could be useful for very remote settlements/islands but even there it probably makes more sense, to just generate the energy locally via wind/solar/geothermal/wave/… as some local batteries would be required in both scenarios.