Exactly, things are changing with V3 to improve reliability. However even with V2 various factors were influencing payouts including bandwidth and uptime so it is not something new. To be able to compete with mainstream storage providers storage node operators need to also be reliable and in my opinion that is for the best. You shouldn't think that just because someone has a spare HD they can put it towards Storj and be able to make a profit. But don't let that discourage you, fill in the form as best as you can and see if you get an invite when the time comes.
It was advertised and promoted literary as "you have free space on your laptop's harddrive, why don't you sell it and make profit!" You can watch some of the videos of Shawn Wilkinson. This is exactly what he promotes. Now it turns out, that to compete with the storage providers, you have to either come up with a miraculous algorithm which negates all latencies combined: from network lag, time difference, dropped packets, disk performance. Or hire some retired admin with old storage servers in his basement (from the above mentioned storage providers) fill them up with drives and waste a few kW of electricity for pennies.
At the same time these enterprise storage providers have gone all NVME Flash.
Storj team have to very carefully reconsider their path, becase there is zero chance to compete with enterprise storage, using someone's kitchen laptop.
I think that the only chance they have is to be a cheap alternative to Tape backup storage. Where you hold the data for a long time, but rarely upload-download it. That might work.
Another path is to convince people that decentralized storage is much safer and more secure than traditional, but that is counterintuitive and will not be easy. And then bet on security.
good luck