While the market for mixed signal design is large and growing the same cannot be said of the EDA tools. Though there is large scope for addressing the challenges in this domain the total market value for tools is not going to be as big as either analog or digital alone and once the interface modeling is conquered it's a plateau.
Yes, the design teams are going to be pressed both turnaround-time-wise and cost-wise. With the design teams getting ever smaller, and the time-to-market pressures increasing, it is going to be a tsunami on the design engineers.....! There is an un-addressed need for EDA tools to evolve.
Instead of true mixed-signal designs, what most fab-less companies do today is to get the analog blocks designed by some external 3rd party and insert the blocks inside digital fabrics. Here the interfacing becomes an error-prone domain. Server processing resources for system-wide simulations and for timing extraction are areas where the lacks in EDA capabilities become pressing issues.
Further, one reason why true mixed-signal design expertise is not abundant, is the lack of widely used and popular EDA tools. Design engineers tend to develop to become either analog experts or digital experts, but not mixed-signal experts. If and when a strong EDA tool emerge, and it becomes widely used, it will help create its own market demand by assisting the growth of a new breed of true mixed-signal design engineers. Therefore, the first mixed-signal EDA tool to gain strong market penetration, may entertain aspirations of becoming monopolistic-ally large.
The wall between analog and digital design has been lowered somewhat, but it is still there and the EDA industry has still not adequately addressed the integration of the two. There are indeed opportunities for EDA companies to "catch the wave," and I sincerely hope they do -- otherwise we designers are going to be overwhelmed by that tsunami!