s0=0.3, si+1=(-0.5+0.25i)*si2+(1.2+0.5i)
Center:0+0i; Zoom = 0.5
Some more playing around with complex parameters in the a*si2+b sequence generator.
I have not investigated the long term behavior of this sequence. Given the large capture set and Mandlebrot-like right side, I suspect it the sequence converges to very short cycle, most likely a single point.
A couple of zooms to follow.

