The problems, according to them (and I have no reason not to believe them), is that right now the burden of proof of the obviousness of a patent application is

The problems, according to them (and I have no reason not to believe them), is that right now the burden of proof of the obviousness of a patent application is on the patent office, not the other way around. And that this obviousness can only be proved by either citing an older patent or a certain type of publication, or a combination of those, not what the examiner "just knows" - or what it looks obvious to him/her or any other technically skilled practitioner, as the basic law says. According to them this "presumption of patentability" (as I would call it) was a policy change that happened not very long ago, as before the policy was to reject as many patents as possible. I would also make a wild guess: That this change was lobbied for by the IP law firms, because they are the only ones which really benefit from it.