As far as I know forks happen when blocks are generated from different sorces and those new blocks aren't make it to the whole network in time. And the fact that we fragment the network by only using connect/maxconnection and refusing other nodes to connect to us with listen=0 I don't think we will ever solve it this way.
With that said I think the best course of action would be if all of us would start generating blocks on the same node, meaning we should all mine on the same pool until we can interconnect all of our nodes. A pool that generates most of the blocks can't be thrown off by forked nodes as far as I know - unless something is really messed up. We should also only connect our wallets to the pool in order to make sure our wallets always get the newest blocks. For miners the dedicated nodes like the three mooo nodes doesn't really help with the forks because they aren't generating blocks just helping the communication between nodes but they do let you fork and ban you if you do because of solomining.
Since the cryptoworld pool solved 504 blocks in the last 24 hours it seems to be the biggest pool so now I'm only connected to it (connect=amber.thecryptoworld.org:4034) and will only mine there instead of solomining, BUT will re-enable PoS so if I get forked it will be safe to say it's because of PoS.
there are as well forks happening by POS...
...and the pool @ thecryptoworld might have been forked as well... but it might as well be me, the blockexperts explorer and mooo1
and the thecryptoworld pool was the only pool being active for the last like 10 hours straigt alone and even longer in accompanyment with zpool when that pool forked eventually... and the hashrate was pretty constant... there were spikes of enormous POW hashrate in comparison but the network seemed to go through those events without any problem... which imho are all evidence contra POW being the issue here...
so, I'd really advocate for really finding the problem in the first place instead of switching to POS only just to 'try'...