FWIW, supposedly the Super Silver Haze is itself a variant of Jack Herer. Basically its just a different selection using the same parent lines.
Sannie offers a Jack Herer x Blueberry called, appropriately enough, "Jackberry". Its not quite stable yet, but he is on an F3 that offers two phenotypes, and reviews are good. The Sannie's Jack parent, is, by all accounts, pretty good, and I also believe he is using original Blueberry genetics directly from DJ short. The point is, if you're after a SSHxblueberry, this ought to be a pretty good rendition of something like that, that you could have sent right to your door in seed form at reasonable cost.
To answer the original question, with the usual caveats that I'm not a breeder, and lack the time space energy and dedication to actually pull these things off, *IF* those things were true, this is what I'd try:
1. Take a lowryder auto, then cross and backcross it to a herijuana plant to create a stable, inbred autoflowering "Hairy-ryder" that was otherwise identical to a conventional herijuana.
Its basically what DNA genetics has already done with its "60-day Wonder", only with Herijuana instead of William's wonder, and (unlike DNA) the backcrosses continued until the plant is absolutely stable. This is actually a very conceptually simple project, and in fact all of the crosses can be planned in advance with only selection for gender and autoflowering trait being necessary. Unfortunately, its still a 2-3 year project that requires the ability to select from at least ten plants at a time.
2. Next one is a little more crazy. The conceptual goal is to create something totally unique, specifically the ultimate "stealth" outdoor strain, using a variety of non-standard appearing strains hybridized together so that the offspring is short, stealthy, autoflowering, rapidly maturing, and looks as little as possible like a conventional cannabis plant.
Using only genetics that I'm aware of, I'd combine an inbred purple indica, a fast/potent ruderalis-based autoflower, Australian 'ducksfoot', and Oaxacan highland "creeper" strains, with the idea of creating a short autoflowering vine-like plant with non-conventional purple "ducky" leaves.
This one is amenable to more conventional selective breeding. EG you could do [(Purp x lowryder)x(Duck x Oa)] x [(Purp x lowryder)x(Duck x Oa)] to create a veritable "explosion" of phenotypes. Then you'd have to do some pretty careful selection to retain the traits you want, followed by several more generations of stabilization. Unlike the above project, this one would be more of a "crap shoot" in terms of finding specific offspring with good traits, and you're really need to generate dozens, if not literally hundreds of plants to do the best selection.
Again, not only is this not really a practical project for an amateur, but I'm not even sure that all these traits can be stabilized together. For example, the Ducksfoot trait, supposedly, is pretty hard to work with. I've never actually seen a "creeper" plant, but supposedly they do exist. Even so, the 'creeper' trait may simply not be compatible with the goal of a short fast-maturing plant. In theory, of course I'd be open to other potential "stealth" options involving non-standard cannabis traits, but in practice, since its just a pipe dream and I'm never going to actually try this, I'm not worried too much about it!
