Your idea discounts the effect of lead platform and lack of incentive of devs to invest in the extra time in platform specific enhancements for a cross platform game without the benefit of delayed release. Producing better PQ on the Xbox and GC versus PS2 was relatively easy as it was evident from day one yet numerous cross platform titles did little to expound the greater capacity of the technology presented by the xbox or the GC. There were popular cross platform games that showed little to no visual improvement when comparing a PS2 game to an Xbox or GC game.
If the current conditions showed a visual disparity between the 360 and the PS3, then that would warrant extra investment as PS3 cross-platform games would have to compete with PS3 exclusive games on a visual basis. However, those condition don't exist and even though heavily invested platform specific games such as HS, F13, Liar and MGS show wonderful graphics they don't show a big enough visual disparity to prove the PS3 technical superiority. Its hard to imagine that cross platform titles would spear head such a movement even with better tools.
That's the thing though, what's being discussed here are tools Sony is developing such that they can hand them off to devs and basically say:
why don't you use this, it's painless and will improve your game. Case in point,
SPURS, that multi-platform devs seem pretty warm to these days with its new availability to them. The tools available go beyond that as well of course, and it's not all about graphics and PQ alone.
Your Xbox/PS2 example isn't perfect either because although you mention that XBox games really didn't get any extra loving, the truth is that as the gen went on, devs
did go ahead and bother to achieve a better game, since the requirements for taking advatange of the memory - for instance - were so so easy to do.
Now I'm not saying a dev house will change the fundamental game mechanics to better suit the PS3, but then again I *am* saying that nor will dev houses be so reluctant as to not just simply turn the key when a turnkey solution presents itself. Remember that XBox and PS2 were fundamentaly different platforms as well; PS3 and 360 are both on the Power CPU/DX9 paradigm, so development flexibility between the two is increased via their commonalities.
That doesn't seem to make sense. A good cross platform engine will never totally maximize the strength of one individual platform. Sacrifices have to be made somewhere if one is to make an engine thats conducive to all platforms that have different technical philosophy. If Dirt performs well on the 360 in comparsion to the PS3 version then either PS3 Dirt doesn't make full use of the SPEs or the SPEs don't provide the type of performance one would expect as one would expect that being the lead platform automatically and inherently would give an edge in terms of quality to the PS3 version.
I agree that a multiplatform game with an engine tuned to a certain architecture over another is not going to max these things out; such of course is the premise of the entire thread. But just as Framework favors PC/360, and PSSG favors PS3, that doesn't mean that a great game for both platforms cannot be created. In DIRT's case, I think that they used the tools to exploit the development areas those tools emphasize, and to the extent that PS3 might have been able to take it further in certain cases, probably just scaled it to a uniform cross-console experience. Who knows - not us until the game comes out (or I interview them), but the tools themselves definitely do
allow for 'more' on the PS3 (in that the Cell is well-suited to the aforementioned tasks).
