Hmm. That’s a great nerd-snipe. (I trust no cite is needed amongst this audience).
Clearly you could build an Earth-sized spherical bag made of [whatever] in free space then fill it with gas ever so slightly pressurized above the vacuum outside such that it holds its shape.
For teeny enough overpressure you could make the bag very thin, lightweight, and flexible like the Mylar used for very high altitude research balloons. Such that a human (in a space suit) armed with a pole vault pole, a javelin, or a sewing needle could poke a hole in it. If they had a way to anchor themselves to the surface; an Earth-sized bag of e.g. hydrogen would have very very little gravity, and you can only poke the surface as hard as you’re attached to it by any/all means.
But due to the teeny overpressure and lack of stretch in the bag material the results would not be very exciting. Some wafting of gas; not a toy balloon-style wild spurting ride around the solar system.
To get the desired wild ride we need an elastic envelope under significant tension and / or / equivalently a significant gas overpressure versus the external vacuum of space. Which quickly gets us into practical discussions about materials, thicknesses, and all kinds of stuff we (well I) can’t arm wave to an answer.
I’ll WAG that any elasticity, strength, thickness, and overpressure enough for a good wild ride around the Solar System will not be breachable by a mere unpowered human with unpowered tools.
Kinda and kinda not IMO. You do raise a good point I hadn’t considered.
Imagine a very thin very stiff shell of superheavy unobtainium. Yes, there might be a lot of gravity there. But if the whole thing is strong enough to withstand its own gravity, it’s not going to shrink when the gas is let out. In more formal terms: lotta stress, very little strain.
Now buckling failure is real, and if the stuff is heavy but weak (e.g. lead), your hoped-for rigid sphere held up by gas pressure will quickly collapse into a randomly crumpled heap once the gas pressure starts relieving through a hole and gravity begins taking its course across an expanse of material that can’t possibly be perfectly homogeneous. IMO the end-state total enclosed volume of the crumpled mess will still be a decent fraction of the spherical whole. So not a great way to provide expulsion pressure, but some. As you say.
Now to the degree the lead envelope starts tearing and leaking gas from multiple orifices while buckling the wild ride effect will be greatly reduced. The optimal ride obtains when the bag contracts isotropically and uniformly while the gas exhausts through one orifice.
