Is there also a Gentoo profile for a system-wide clang/llvm? Add libressl and musl (instead of glibc), and it would be quite a different and advantageous linux distro.
It's awesome that cool things from OpenBSD are being ported over to FreeBSD, but why not just use OpenBSD from the get-go? It's already a struggle having to deal with FreeBSD's outdated version of pf.
I've got a lot of love for OpenBSD but I can understand people wanting to use FreeBSD for things like ZFS and dtrace. However for those seeking to use said features personally I would recommend looking at an illumos distro.
There also BoringSSL[1] but that might be even more of a departure than LibreSSL in terms of API compatibility. I still think it's surprising that we don't see more BoringSSL being used especially with nginx.
There are some good reasons not to use it. Primarily the lack of API stability and that, as you mention, the people making it caution against it.
But to answer your question, why would someone choose to use it anyway? One reasonable justification is that many people believe (probably myself included) that the quality is superior to openssl or libressl. The APIs are unstable but the flip side of that coin is that they're probably better. Also the engineering practices behind boringssl have lead to what I would call relatively high quality code. It's well structured, clear, and maybe less likely to suffer as many serious bugs as the alternatives. Time will tell.
Not that I'm recommending it for everyone, just answering your question. Quality is often in the eye of the beholder, use what works best for you.
This really depends on what your production environment requires. LibreSSL is mostly API compatible with OpenSSL but removes FIPS and support for esoteric platforms. If your production environment requires Windows 3.1 or big endian amd64 then probably LibreSSL is not ready for your production environment. Seriously though some of us are already using LibreSSL in production without problems.
This really is a more complex question then it may seem. One of the reasons I haven't yet upgraded my production system is because I want to be able to keep my system up-to-date. Now, having to manually patch the FreeBSD source tree once a new upgrade lands is a bit of a pain. The point of the article is really about exploring where we are and where the FreeBSD community is headed.