The GPU in the PS3 isn't as powerful as the Xbox 360's GPU, but there are a number of great libraries available that permit augmentation of the GPU with the SPUs.
From experience, architectures that are designed first for the PS3 run very well on the Xbox 360, but architectures that are designed for the Xbox 360 (or, worse, PC) will not perform well on the PS3, which punishes you for pointer chasing. The tile-based lighting in BF3 that you mention is interesting because the Xbox 360 CPU did not have enough horsepower to crunch the lighting data so they had to leverage the GPU to augment it. This was challenging because the Xbox 360 GPU was a DX9.0c+ era part and wasn't intended for GPGPU computation. Colin Barre-Brisebois spoke at SIGGRAPH about how this was done: http://publications.dice.se/attachments/BF3_NFS_WhiteBarreBr...
I really like developing for the PS3. It's challenging to code for initially but the rewards are definitely there if you try.
Seeing that you edited with a citation from a Volition AP. I guess if you trust sources like this - there is no argument possible.
>In some extremely rare cases, some game engines found a way to use them for graphics (like Battlefield 3's Frostbite 2 engine towards the deferred lighting passes). But by the time engineers found a way to leverage this extremely complex architecture, it was already too late.
You mean like 2007 Uncharted?
The PS3's internal architecture was poorly balanced. This really hasn't been a secret for a long time, for anyone who can take off the hype goggles. It sounds like the PS4 will not have this problem. It will be interesting to see if Microsoft makes the same mistake; it really won't be a console-gaming win to stick 8GB of slower RAM in the box, then suck away 2 or 3 GB for other purposes. (Probably still won't be as poorly balanced as the PS3 though.)
But look at something like Halo4. It's had all of Microsoft's drive behind it. Pretty, yes. Significantly better than anything the PS3 could do? Not really, no.
And I agree, PS3 was not a nice architecture compared to 360. PS2 was not nice compared to xbox 1 too. However this just not justify idiotic claims like "nobody used SPU" etc.
Almost nobody used "cell processors" to potential initially, and for very good reason [2]. The bottom line is they were incredibly difficult to program for in any useful way to graphics. But more than that, they are limited in very specific ways [3]. They won't increase triangle throughput, and they don't directly increase pixel throughput.
In some extremely rare cases, some game engines found a way to use them for graphics (like Battlefield 3's Frostbite 2 engine towards the deferred lighting passes). But by the time engineers found a way to leverage this extremely complex architecture, it was already too late. If you actually look at most Xbox360 vs PS3 games side by side, the Xbox360 often looks much better.
In this case though (the next gen Xbox and Playstation) the situation is reversed: Sony's GPU is actually legitimately faster, and by a large margin. Not only that, but having a full 4GB of fast ram is a HUGE advantage to Sony. High performance 3D rendering is inherently a bandwidth hog, because everything on the screen must be pushed through the pipes every single frame. Having a fast 32MB cache doesn't really help you if you want to consistently and smoothly render more than 32MB of content on the screen. And moreover, just as the last generation PS3 had an overly complex architecture making it difficult to reach its full potential, it seems Xbox and Playstation are swapping roles here as well with this generation.
[1] http://forums.gametrailers.com/viewtopic.php?f=23&t=7947...
[2] http://www.videogamer.com/ps3/saints_row_2/news/two_or_three...
[3] http://stackoverflow.com/questions/1355827/what-does-program...