That is an excellent question. There is really a lot of cross-bleed between C++11 and D. In fact if not for anything else D was good competition to wake the C++ giant up into doing something about its deficiencies and backport as many D features as possible. I think this will continue to happen and and I think both the languages will benefit technically from this dynamics.
As for, should one abandon and existing C++ project and rewrite in D. Probably a bad idea, such a thing would require strong arguments. On the other hand for some project that is beginning now, I would say its evenly matched. The possibility of abandoning a lot of the C++ cruft and legacy is not something that should be ignored. Whether that compensates for the reduced maturity of tooling as compared to C++ has to be decided on a case by case basis, and this question is going to be a perennial question that would afflict the adoption of any new language. Its hard to have authoritative answers for this. That said I am looking forward to some improvements in D, better garbage collection, better separation of functions in the std library that uses garbage collection, getting the runtime memory requirements down for compiling D code that uses a lot of CTFE, and well if you could add sum types and pattern matching and efficient fibres/coroutines that would be very nice. I think fibres would be a tough one if one has to maintain portability and compatibility with C libraries.
As for, should one abandon and existing C++ project and rewrite in D. Probably a bad idea, such a thing would require strong arguments. On the other hand for some project that is beginning now, I would say its evenly matched. The possibility of abandoning a lot of the C++ cruft and legacy is not something that should be ignored. Whether that compensates for the reduced maturity of tooling as compared to C++ has to be decided on a case by case basis, and this question is going to be a perennial question that would afflict the adoption of any new language. Its hard to have authoritative answers for this. That said I am looking forward to some improvements in D, better garbage collection, better separation of functions in the std library that uses garbage collection, getting the runtime memory requirements down for compiling D code that uses a lot of CTFE, and well if you could add sum types and pattern matching and efficient fibres/coroutines that would be very nice. I think fibres would be a tough one if one has to maintain portability and compatibility with C libraries.