I've worked for a company a few years ago that during one period had a small (3-4 person) true UX team for a large development crew (had something like 160 java devs all in same building). They produced the wireframes based on client services' input, tested with users in a controlled environment, did the heatmaps, produced the wireframes for the dev team, etc. However, the company axed it along with the initiative to centralize core functionality. Why? The primary reason was that it was hard to determine how it was helping the bottom line and the centralization and addition concern over UX was slowing down the overall pace. I still think that people need to spend more time thinking about UE/UX and spending time improving the UI based on real world data. However, what will they axe money for first when the s--- hits the fan?