They can start small and experiment. Take a tile, do a one hour window at 10am, and then a 1 hour window at 6pm and go from there. It'll be very close to a normal release, but expanding it a bit. Then compare the data against what they likely would have ordered normally and see how it goes. Do this for a couple titles, to get a better sample and see what happens. The negative impact on their business should be minimal in the worst case scenario, but the potential gain and information learned would be huge and worth whatever potential dip that could happen. I'd say it's no riskier than increasing the number of titles they're doing because when you open up to more titles, you risk some games not being as popular too. That's just one thing they can do in a relatively low risk manner.
Other things they can improve is purely their communication. They're so bad on it a lot of the time and it's been like that for awhile now. Their business is old enough now that this should be something that at the very least is getting better if not fixed by now. A lot of the frustration people have is partly due to their communication as well as the feeling that they've been doing this for awhile while still making similar mistakes.
Heck, maybe someone else here can elaborate why Shopify is better than all the other storefront alternatives too. There are plenty of options out there, so what does Shopify allow or do better that the competition doesn't. If nothing, why isn't LRG considering changing? It would be helpful to know why we're stuck with Shopify which is clearly causing problems. Better communication about this would at the very least help people understand why they're getting screwed rather than just hand waving it away as that's the way Shopify is.
There are so many things LRG can do to improve things. Some are way easier than others and while this was okay early on when they first started out, growing, and learning, it's starting to become tiresome that it feels like not a lot of progress has been made to improve these things now that we're what 75 releases in? When a game sells out in just a minute or two, and you continue to do that, you're doing something wrong by being way too conservative in your business. While it's understandable that some games might be harder to predict, they are also getting some bigger titles that should be far more easier to also predict. Night Trap, Wonder Boy and Ys are prime examples of these.