No, streebog is much slower, infact it's similar to groestl but much heavier. Unless you can bitslice streebog LoL!
(Here is a Profiling of the SIB opensource kernel by the army-boy on the gtx 1070)
Streeborg+keccak 29.2% of the time.
Groestl 19.02%
With my 5% faster groestl, I get a 1% faster SIB kernel.
==4580== Profiling result:Time(%) Time Calls Avg Min Max Name
29.20% 4.84423s 53 91.401ms 89.100ms 93.843ms keccak_streebog_gpu_hash_64(unsigned int, unsigned __int64*)
19.02% 3.15548s 53 59.537ms 58.116ms 60.683ms x11_simd512_gpu_compress_64_maxwell_echo512_final(unsigned int const *, uint4 const *, unsigned int*, unsigned __int64)
16.07% 2.66591s 53 50.300ms 48.746ms 55.073ms quark_groestl512_gpu_hash_64_quad(unsigned int, unsigned int*, unsigned int*)
14.67% 2.43401s 53 45.925ms 43.805ms 48.247ms x11_cubehashShavite512_gpu_hash_64(unsigned int, unsigned int*)
7.37% 1.22197s 53 23.056ms 22.444ms 24.419ms x11_simd512_gpu_expand_64(unsigned int, unsigned int const *, uint4*)
4.87% 807.63ms 53 15.238ms 14.901ms 16.200ms quark_jh512_gpu_hash_64(unsigned int, unsigned int*, unsigned int const *)
4.33% 718.82ms 53 13.563ms 12.745ms 14.122ms x11_luffa512_gpu_hash_64(unsigned int, unsigned int*)
1.73% 286.61ms 53 5.4077ms 5.2447ms 8.8874ms quark_skein512_gpu_hash_64(unsigned int, unsigned __int64*, unsigned int const *)
1.69% 280.45ms 53 5.2915ms 5.0875ms 5.7360ms quark_bmw512_gpu_hash_64(unsigned int, unsigned __int64*, unsigned int const *)
1.06% 176.02ms 53 3.3211ms 3.2270ms 3.6693ms quark_blake512_gpu_hash_80(unsigned int, unsigned int, uint28*)