I'm certain that your teams working on this were both tallented and put the best available effort given priorities to doing a good job.
Neither I, nor anyone else, can do a good job evaluating the difficulty and probability of sucess for RC and PR for your existing chips and software architecture, without access to the design information you hold locked up. So, it's impossible for anyone outside your NDA circle to have done their homework.
I've also spent 35 years walking into the middle of clients projects, which were frequently stalled, failed, and/or past contract ship/delivery dates. Nearly everyone of those teams were competent, and many top in their field. They had also reached the limits of their formal training, skills, and experience to find a solution to their project deadlock -- OR -- that were locked into failure by the product specifications forced on them, resources made available, or restrictions against using viable alternative designs, architectures, etc. Frequently the solution path for the projects was a combination of outside ideas, outside experience, and outside influence to change the product specifications, resource alloctions, and removing the road blocks to other viable solution strategies.
I'm certain that if it was easy for you to do with the requirements, resources, and restrictions place on your developers, that you would have delivered a strong reliable RC and PR tools by now.
I'm also certain that ourside your organization, free of the requirements, resource limitations, and organizational restrictions that a different team will respond to a different requirement set and find a viable solution limited only by the existing hardware architecture. And with that success, they will be able to clearly articulate the changes needed to make future Xilinx RC and PR product generations not only very viable, but a strong success for everyone involved, including Xilinx and your customers.