This paper asserts something innacurate: entropy pools can be "used". Entropy is not an object, it's an estimation. Just as you don't use temperature, or barome

This paper asserts something innacurate: entropy pools can be "used". Entropy is not an object, it's an estimation. Just as you don't use temperature, or barometric pressure, you don't "use" entropy. Further, once a CSPRNG is properly seeded, there is no need to concern yourself with whether or not it can produce "high quality random numbers", provided the cryptographic primitive behind the CSPRNG contains conservative security margins. The Linux kernel CSPRNG uses SHA-1 as the underlying primitive. While SHA-1 is showing weaknesses in blind collision attacks, it still contains wide security margins for preimage attacks, which is what would need to be applied to a key generated by a CSPRNG (you have the output, now predict the state that produced it). Even MD5 remains secure against preimage and second preimage attacks. Again, once properly seeded, the Linux CSPRNG can produce data indistinguishable from true random indefinitely until SHA-1 is sufficiently broken.