I don't think this really has to do with open source vs. closed source. But let's look at what would happen if you were paying for a product, and were given a binary blob in order to interact with it. You find a bug. You report the bug to the company. They triage it, and decide it's an edge case, fixing it isn't a high priority for them, and they toss is somewhere in the backlog and expect to get to it in 6 or 12 months or whatever.
As a customer, you're stuck. Either you figure out a workaround on your end, live with the bug, or find a new product to use. The last option might be difficult, since switching costs are often not small.
If the client was open source, you could end up with the same reaction from the company, but in this case you have the ability to try to fix the problem yourself. Should you have to, in an ideal world? No, of course not. But we often don't live in an ideal world.
I agree that a company shouldn't have the attitude that open sourcing their client absolves them of all responsibility for it. I mean, sure, if they want to do that, that's fine. I would certainly guess that would make their product less attractive to many potential customers, but if they're ok with that, that's fine. It's up to the company to decide where to spend their finite resources, and a model where they let their users fix their bugs might actually work out best for them, even if it doesn't for everybody.