There are a lot of great things about it.
But I think they can't be taken 100% seriously unless they take some steps to improve their image.
Basic stuff to make them seem like a legitimate journal, not something someone did over spring break.
I would recommend the following 7 things:
1. Make the website a lot faster.
2. Hire someone to organize it logically. It's nearly impossibly to find things one wants. I'm thinking they could use the Google+ system, since no one else is using it. That way they'd have a search function that works. But also, they need to hire someone who has a sense of website navigation. It's almost comical how hard it is to find the thing you want.
3. Avoid mission creep. As Frontiers keeps expanding, it seems more and more like a Ponzi Scheme, expanding into new areas as old ones wither. This would also reduce clutter on the website and make it easier to find things.
4. And on the mission creep domain, Stop trying to be Facebook. Scientists already have a social network. It's called Facebook.
5. Stop spamming reviewers all the time. Especially with articles that are distant from anything we know about.
6. Give up on the 'interactive review' theory. Most people I have talked to informally write a normal review, put all the comments in one box, and ignore the rest. And then get really annoyed with more than one extra round. We are giving out time to review. We don't want to have a conversation.
7. I dont know if it's just me but there seem to be cookie problems. Frontiers never remembers that I am logged in, and asks for my password every time I go there. No idea why. I dont have this problem with most other websites I use.