I think you're making an incorrect assumption here. If the sol-ark, and charge controller aren't doing closed loop communications, there is no way for the sol-ark to actively limit the amount of power generated by the CC. If there is "power to consume" and enough available solar array to provide the power, the CC will produce it.
I suspect the problem is the opposite of your assumption. The sol-ark doesn't *know* about the additional current/capacity, and thus can't make intelligent decisions on how much power to send back to the grid based on the "excess". The sol-ark is likely balancing "available power" and "power to deliver to the grid", completely unaware that there is something else adding power on the DC bus.
With that in mind, I think the sol-ark is only "limiting" your CC output because it's only selling back the power it knows about. And because it's doing a good job of it, there is nowhere for the "excess" power to go from the CC.
So, as others have suggested, you can change the settings to tell the sol-ark to deliver "slightly more" to the grid during a ToU window, or you can increase the loads on the protected loads side.
Obviously, you can throw more money at the problem by buying hardware to AC couple it, but for 2kw, is it worth it when the other options are free? That answer depends entirely on your situation. If the other arrays cover the majority of your bill, then I suspect even the AC coupled capacity would be "wasted". The other side is, what is the cost to buy the microinverters or other hardware involved to convert that 2kw to AC coupled.. and the payback time? Is it better to make a couple setting adjustments to utilize as much of it as you can now, or is it better to buy the hardware, and utilize as much of it as possible? The payback period probably would be the deciding factor in my opinion.