While I certainly see why that information (playable settings) would be useful to you, there are many sites that will give you that information.
I only speak up because I truly hope W1z doesn't change his graphs...EVER. One could say 1080p may be more useful than 1920x1200, for instance, but you could just as easily do the math (results*1.11).
The list of cards, static results on a control, and percent performance numbers are invaluable when looking at the them from a raw, scientific point of view. This makes it helpful to not only know where cards should be priced to be a value, but identify architectural bottlenecks and potential product gaps (performance, power use that is obviously dictated by pci-e spec, markets based on what is needed to avg playable frames in a popular game etc). They are a one-of-a-kind resource.
For instance, we all expect Hainan to launch at some point. People like me would say Bonaire was reviewed the third week in March and launched beginning of April, Malta is getting reviewed the third week in April, there is a share holders meeting on May 16 (end of second week in May), Computex in the first week of June...
So obviously it's launching in Q4.
I digress.
While you can assume it will perform better than 7870XT/660ti, these greatly help you put expectations in check as well as anticipate realistic pricing for such products, as well as where competing products may land as typically pricing does level out to be in-line with performance relative to each companies' stack...even if we ignore where AMD tried to force pricing 660ti/670 with TahitiLE/7950 pricing.
In short...please don't go pushing for W1z to change his tests. Nuts like me love them for theorizing and planning things out (like potential Haswell upgrades in June etc), and they're the defacto source for people to prove a point when comparing cards or helping others make a choice.
As for Bonaire, I think the purpose of this chip is clear (gk106 will never touch it's potential price because of die size, even salvaged, while maintaining a useful advantage over other options because of performance and efficiency). I still think it needs a config with more memory bandwidth though. Most people are not going to achieve almost 7760mhz overclocks with 6000mhz ram...although if they could 7790 would be perfect as-is at it's inevitable similarly-cheap-to-650ti-but-always-cheaper-than-ti-boost price (almost certainly a theme that will continue going forward; better than equal-priced nvidia part, cheaper than part with more overall potential even if similar at stock). But hey...there's a whole lotta time til 20nm...who knows. I just want what's happening here to happen in the market that's twice the spec and twice the price but smaller than Steven Tyler's...record of blues.
At least we shouldn't have to wait too long.