Phoronix did a fresh six-way Linux distribution comparison. OpenSUSE tended to be the biggest performance outlier in this testing, likely in part due to sticking to the older GCC 4.8 compiler and XFS+Btrfs file-system combination. See http://www.phoronix.com/scan.php?page=article&item=linux-sep15-6way&num=1
Yup, phoronix. A tiny bit of my ideas on phoronix is already worded in the comments.
Saw this as well. Somewhat embarrassing for OpenSuse fans. You like your distro to be on top… even mid range would be nice… but bottom ranking is hard to take.
Come on, they’re testing unfinished distros. And have presented “results” like this before. IMHO phoronix is as good as all the “what’s the best linux desktop” polls.
A test result where f.e. 1 bug fix would completely change the outcome, is no test result.
To be fair, so are we “testing” the unfinished Leap 42.1, and the threads in our forum are available on internet, are they not?
Did Phoronix hide anything about Leap as a Beta release?
One might hope the Leap makers review the benchmark results as feedback from Beta1, and investigate. 
Seems fair enough; perhaps they could have given more emphasis to that point and how big a change leap will be, but maybe they are assuming that anyone who reads this will understand all of the background.
It isn’t opinion, or even a list of arbitrary people’s opinions, it is a set of measurements. Those two things are quite different, so the argument that one is as good, or as worthwhile, as the other seems a little unclear.
The results are what the results are; no point in saying that a bug might change things, unless you know of this bug and that there is a chance of fixing it before release. if you can point to, eg, an old version of SQLite, which would change many results, then that would be a point worth making, particularly with the amount of time to fix it before final release. But, there seems to be no single app that could do this, because it is widespread across many applications.
On the other hand, they do point out that Leap uses a ‘different’ set up from everything else; I can’t see the older version of compiler causing some of the bigger variations that we see here (and Manjaro shares that, so if that were the key, you’d expect to see a similar pattern with Manjaro), so my feeling is that something about the disk setup is a more likely cause. In fact, I’d go further and suggest that the use of BTRFS is the most likely culprit, as BTRFS is known to produce some variant results like this (not just slower results, but results that are different from the standard ext4 results).
Now, don’t get me wrong, ‘performance’ of this sort is not the only kind of performance; if, for example, you need to roll back to a snapshot, then you’ll be happy to pay a few mS here and a few mS there in order to have that possibility. So, in some use cases, the ‘better performing’ system will be the slower one where you don’t have to go back to the backups to roll the system back to a working state, and BTRFS can give you that possibility (mind you, that’s not the only way of getting that particular advantage, so I am not suggesting that BTRFS is an automatic win).
To me, there are questions about whether the BTRFS/XFS setup is really the correct default (you should certainly be able to select that, but should it be the default? is it the right setup for people who just want a quiet life, and who aren’t really techies?) and whether it is well enough explained what is going on, particularly with regard to the space consumed in snapshots.
No opinion here on XFS, never used it. On BTRFS however, the release manager for openSUSE 13.1 turned it down as too early for default status but 13.2 would be about right. The push was coming from a Btrfs developer, and possibly from SUSE who were developing Snapper for Btrfs mainly, and who also made it their default. The questions are still being asked about the openSUSE default, is anyone listening?
IMO Snapper shouldn’t be the default on the desktop version, it should be for User’s/Admin’s to select at installation time if required. The reasons have been posted around the forum, repeatedly for long enough so I won’t do that here. That leaves the question for Btrfs - need it be the default without Snapper and snapshots? My current answer is probably not, as long as ext4 is available and practicable. There are still questions over Btrfs performance for some disk-intensive applications. 