That is a very odd device. I would not say that 0.9 A is overkill. It may very well be underrated.
This datasheet doesnt tell all if you ave never worked with triacs before. You ave to provide for this device, 25 mA for a minimum load, otherwise the triac will always think it's ON.
Now, I'm not sure exactly what the fancyness is doing, yet, but you can turn on a triac by keeping the LED circuit powered.
But, you CANNOT just turn te triac on an expect the zero crosiing to turn it off, like you might with a light dimmer.
Why? Because the voltage and the current are not in phase. In te simplest case, let's assume that the voltage is sinusoidal and the current is sinusoidal, but not in phase.
So, you turn it on and start counting when the voltage is zero. At some point the current will become zero, but the voltage will not be and the motor will turn off. So, you don't get what you expect.
To pase angle fire any such load, you have to do it periodically or continuously, between the zero voltage crossings. When the device goes through a current =0, it will turn off. With a resistive load V(t),I(t)=0 at t=0, but not for an inductive load.
Now, since you are creating this pulse train for on, you have to monitor the current in the load. Thus if the peak current is exceeded, the pulse should be removed.
One system I worked on was able to limit I this way. Max I becomes a settable paramter. In the systems I used they used back to back SCR's for th AC line. One reason is that both could be fired togeter to mediate surges. The adjustment range was 10%-100%, so a 10 A triac could be limited from 1 to 10 Amps.
During my tenure, I convinced them to always use pase angle fired with current limit and always use 25 A SCR units as they were called for 10-15 Amp loads. These devices had integral semiconductor fuses at $30.00 a pop and they were popping them left and right due to the nature of the system. Adding a properly sized 3AG fuse cut the number of semiconductor fuse blows significantly.
Each of these triac thingy were operating into a transformer to get like 30 V at 10A out.
Later, I conviced the powers that Be, that A DC power supply would save money in the long run, be smaller and would not pop fuses. As an added benefit, the power calculations of V * I would be real.