The trend in the build up of tech supporting the eventual sound update indicates that, for some time, PD know that "sound makes everything". That's likely why they've taken such a risk in re-evaluating the whole process from the ground up.
The gamble, of course, is getting something working and into the public eye ear within a "reasonable" time-frame. The other gamble is whether or not you can actually deliver an improvement with all that effort - my own experiments say yes, certainly in terms of expression.
However, the work they've been doing goes far beyond experimenting with recording tricks or multi-tapped reverb sends, and that's just looking at GT5 Prologue. For instance, the relativistic source mixing is a serious statement of intent in itself - that's something that other developers have tried, but because it immediately leads to more problems (that require more ground-up solutions, and the understanding to go with it), it's quickly abandoned. PD stuck with it, and now it more or less works perfectly.
Again, it's taken them a long time to achieve that, and that's the gamble (one they perhaps under-estimated). In any progressive endeavour, risks have to be taken; otherwise you just end up standing in the same place.