I don't think there's really much energy in ocean currents. Based on the numbers from here [0], it seems you can get no more than 20 GW from the entirety of the Kuroshio current (the big one near Japan that they're targeting). That's not really a lot, and presumably the environmentally safe level is a fraction of that, after which you're fucking up heat transport on a planetary scale*.

(Maximum flow of 65 sverdrups = 6.5e7 m³/s, and a top current speed under 1 m/s. Density of water is 1e3 kg/m³, and there's also a subtle multiplicative factor of 0.6, from Betz's law).

[0] https://tos.org/oceanography/assets/docs/28-4_andres.pdf ("Mean Structure and Variability of the Kuroshio")

https://en.wikipedia.org/wiki/Sverdrup

https://en.wikipedia.org/wiki/Betz%27s_law

*(Footnote because this stood out to me as weird, but the math holds up: the amount of heat energy transported in ocean currents is >10⁴ times greater than their kinetic energy (what ocean turbines can extract). 1 kg of water stores 4,200 J per 1 °C temperature increase, but only 0.5 J of kinetic energy at 1 m/s. So I think that blocking an ocean current would have a disproportionate amplifying effect on the climate. IANAO).