More colloquially: I don't know a single founder or entrepreneurial engineer who thinks "oh great, ERISA, let's learn more about that today!"
Could a firm position itself as having all owners and no employees to avoid these complexities and operate more efficiently on private terms? Has anyone seriously tried this in tech? Outside of tech there seem to be narrow successful efforts (single airline pilot LLCs, gig work, etc).
Presumably, such a firm would have to find analogs for common arrangements such as compensation (salary, equity, benefits), but surely that could be done with a collection of standard private contracts.
To name just a few potentially-not-great but hopefully illustrative ideas:
(a) Instead of hiring a full-timer, perhaps the firm would co-found a joint venture with every engineer or executive. With standard YC-style docs, this could avoid issues of investor accreditation and all-or-nothing access to IP/systems. IP would be cross-licensed between entities on limited terms, systems shared across zero-trust boundaries. Entities could be acquired or merged over time to simplify/cull the structure. Buyouts could be like bonuses - on better tax terms.
(b) Instead of salaries, firms could offer loans to their owners, to be forgiven incrementally in exchange for various services rendered, for instance in the form of contributing IP to the enterprise by committing code or delivering on objectives or providing services. Owner draws and dividends could be alternatives.
(c) Instead of layoffs, companies would simply not renew a contract, exercise termination clauses, rescind a license, or resign from a JV board. This would be massively simpler than RIF constraints, say. Entrepreneurial minor shareholders might face harsher (but negotiable) termination terms; they also might enjoy much larger upside if their firm doesn't have to budget for and overcome the drag of a heavy system.
Labor laws foresee some of this by treating certain situations as de facto employment relationships, even though they are not papered as such (permatemps, independent contractors who work with only one firm, etc). An individual software engineer working with a group like this could be a co-founder in some entities and an independent consultant/contractor/vendor to others, for example.
Clearly this is not for everyone, but self-selection should be an asset. No doubt, these types of arrangements would suffer overheads and non-trivial legal costs of their own. At least this system would be of a firm's own design and would be based on (more predictable/algorithmic, less emotional) contract law, as opposed to the current (more emotionally/morally charged) situation. Standardization could mitigate these overheads and the structure could be a source of competitive advantage if used judiciously by skilled operators.
In the limit, perhaps this type of structure could also offer paths to scalable value transfer between collaborators with options to devolve control/apportion responsibility more gracefully and to concentrate power a bit less in some dimensions.
(I also noticed the word partnership was missing from your analysis but again, don't.)
https://www.usworker.coop/clinic/startups/
I think the difficulty for both those and your proposed structure is the difficulty of fundraising, when outside investors can't just invest capital without labor. There are lots of rich people and institutions with more money than time. And conversely, not every employee/potential co-owner would necessarily want equity in lieu of a stable salary.
It also gets into managerial complexity when co-ownership might also imply some degree of democratic decision-making because everyone is not only a worker but also an owner. It doesn't necessarily mean that everyone owns an equal share, but hiring and firing become more complex, and traditional company hirerachies may not be as common.
There are a few tech coops: