Well, certainly more realistic than living on the surface. And probably easier to set up than a Mars habitat - terrain is irrelevant and your entry is so much easier - plus, even normal Earth air is a lifting gas on Venus. And it'd be no less self-sustaining (that is, to say, "not very"
There's no need to send people offworld to do science, whether to Venus or Mars. But while there's no need for any kind of "facility" at all, manned or otherwise, for robotic equipment on Mars, the concept of some sort of floating "facility" on Venus is pretty important. Any sort of craft designed to tolerate Venus's surface environment is going to make a terrible analysis lab or sample return vehicle. I mean, even solar panels would have to be heavily shielded on a sampling run to not be destroyed; there's very little that you can have exposed that can tolerate that environment. Sampling and analysis or return on Venus is best done in two stages: 1) Buoyant craft that repeatedly dive and rise the atmosphere like submarines and take samples on the surface, and 2) a floating platform containing any analysis equipment or return hardware, high gain communication with Earth, and solar panels to recharge the batteries of the sampling craft while samples are being offloaded.
Venus's surface is really unusual and it'd be neat to know more about what's there. I'm still not big on the concept that we need humans there to do it, but at least a floating platform of some kind would be important. The only advantages I could see for having humans would be to cut the communications latency with the samplers to allow for smarter sampling decisions without requiring them to wait in the harsh environment for round-trip communications on Earth, the ability to repair samplers, and perhaps mildly better local analysis of samples and/or decisions about what to bring back. Hard to justify the added price tag, though.