Claus P. Schnorr is a big name in cryptography, no doubts, but I'd not take this work of him as a serious one for many reasons:
1- It falls on its own because of an irrational conclusion it has made about 800 bits RSA keys being breakable by just 4.8*10
10 arithmetic operations. It is literally nothing, like few seconds for a modern cpu, and there are many 800 bits challenges out there waiting for a champ with a brilliant algorithm to overcome in real time.
2- SVP and lattice based methods in general terms, are not new technologies, they have been around for quite a while, and it is very unlikely to disrupt the integer factorization problem by applying improved versions of such methods.
3- In this paper, Schnorr, again uses his suggested pruning technique from the 1990s, which is not established as being rigorous enough.
4- Even in its first stages of getting peer reviewed, the paper has received
strong backlash from cryptanalysts.
Conclusion
No, RSA, the way it is used in the industry with very long keys, is not close to a break point, even for 800 bits long keys which are obsolete anyway, state-of-the-art algorithms can not do the job by utilizing less than two thousands CPU-core*years.
No need to remind that the whole story has nothing to do with bitcoin as RSA is not employed here.
EDIT:
In spite of the last point I made above being absolutely valid, it is also true that such a hypothetical breakthrough in the integer factorization would be somehow unpleasant news for Bitcoin because it suggests the feasibility of similar developments in the discrete logarithm problem field which ECDSA is based on it.
