This is the so-called "binding problem". I found an elegant theoretical solution some years back, based on resolving informational and connective entropies in signal ana|ysis and network equilibria, as determined by conservative selection pressure and cellular geometries (ie. thermogeometric ambient potentials and complexity, or rather, simplicity, in randomly self-organising networks). Basically, efficiency is 'god', quantum information is a four-dimensional state space (partial, prior, temporal and spectral), and all information we can see, hear, taste, smell, think or feel is PPTS modulation of factor-two synchrony. Which was nice.
It's too much to post here, and I don't have a write-up online, but see Miller et al's work on temporal thalamacortical modulation:
http://jn.physiology.org/cgi/content/full/87/1/516
GC Huth's pioneering work on the retinal transform:
http://www.ghuth.com/vision/
Also Axel and Buck's 2004 Physiology Nobel for their explanation of olfactory function, and subsequent papers on this and gustation, David Mckay's work on cellular function and information theory, as well as anything on set theory, Fourier & wavelet transforms, just about any of the work being done at Penn, Stuart Kauffman's work on dynamical systems theory, anything on synaesthesia, and just about anyhting else that seems even remotley connected. I went through some three gigabytes of PDFs in one year, trying to pan it out. Did my ' head in, but then I ain't not no acedemic nor nuffink, innit...
But I think you're right - We'll get there, eventually. kurzweil's singularity and all dat, eh. If the 'terrorists' don't take us all out, Skynet probably will. Hope I get to play E4 beforehand...