Correct. Unfortunately systemd is not a well separated 'stack' in the middle between kernel and applications but an interweaved network which tightly connects all three parts.
I hope that the community has learned from the SystemV/systemd discussion and will go a new way which avoids the weaknesses of both systems towards a cleanly separated layer.
I'm sure the BSD community can learn and can improve on the idea. Perhaps their result could be ported back to Linux as an alternative.
EDIT: Toned down possibly inflammatory language
This is an infamous example: http://lwn.net/Articles/430598/ [gets particularly heated at mezcalero's top level comment]
Linux users make jokes at Windows' expense all the time. It's usually not higher profile developers doing it, but nor have they ever, really.
And he clearly has a point; at its worst it is a giant ball of YAGNI.
Do you think that Thompson and Ritchie gave a !%^$@ about Multics compatibility? Or OS/360?
Free software doesn't need to even support free operating systems, for that matter. I'm sure that there are Windows-only apps that are F/OSS, though I don't know of any off-hand. Of course the systemd developers don't need to support any other OS to write "Free Desktop" software, any more then the FreeBSD developers shouldn't feel compelled to port their filesystems, init-system, or anything else to Linux.
Was this agnosticism part of the design?
And even if it was -- I suspect his experience has altered his view on it.
Rather, it's his patently false and almost obnoxiously smug view of the BSDs as these slow-moving operating systems that simply follow Linux's trail, and his deeper implications of forking being negative and that Linux should play a game of catching up against OS X and Windows instead of doing its own thing.
Then his reasoning against portability seems self-defeating. He says that "i need to reach customers using windows" is a good reason, in which case you can substitute "windows" with any OS and have a purportedly valid reason for portability.
Not living in an arrogant monoculture that seriously believes you're leading the pack and everyone else is biting your dust sounds like another good reason.
This is a nonsense false dichotomy.
Just saying what you do well will result in your competitors saying, "we do that well too", and won't win you any market share. You have to demonstrate that, "no, they really don't do that well." So it's a real balancing act between promoting yourself and not being a dick to your competitors. In my own personal case, I screwed that balance up royally, and it's a problem that has followed me around for the last decade.
Similarly, look how hostile OpenBSD is toward FreeBSD about security designs (watch any Theo de Raadt presentation.) I almost never see any FreeBSD users taking pot shots at OpenBSD like that. Then look at the market share of both, and it makes more sense.
Well, watch the interview with Shawn Webb on BSD Now #52 http://www.bsdnow.tv/episodes/2014_08_27-reverse_takeover and think about his comments on OpenBSD. A lot of NIH in that interview. I don't think it is isolated to one side as you think.
Then the more logical explanation is that FreeBSD is lagging behind exploit mitigations.
Keep in mind that Theo actually praises Windows and says that they're second after OpenBSD. If this was about being envious over market share, then that wouldn't make much sense.