That being said, if you have the knowledge-working talent, organizational skills, ambition, and reputation necessary to start a successful guild, you can make a lot more money as a consultant, and money is a powerful motivator.
I think if, as an economy, we were geared more towards quality and building for the long-term, a software guild would emerge more naturally. We've optimized for a different system, that doesn't much reward the differential outputs of a guild-like association. Reasonable people can disagree on the relative merits of this system. :)
When you hired somebody from them, you knew you were getting the real deal. Their contracts included protection from off-the-clock hours, 4-day weeks, etc. They were also super transparent and accurate with the "level" of people you were getting. They charged a premium-level fee (read: not an exorbitant, luxury-level fee), but my company at the time was happy to pay – the marginal expense was basically insurance that nobody involved was going to be wasting their time. On every IC team they were embedded in, they got better performance reviews and 360 reviews than anybody else.
I think over the years, they (tragically, IMO) appear to have drifted away from this model, because they found they could grow faster by aiming more "mid-market"; my guess is that the pool of money on the buy side in the mid-market is larger.
Basically, mediocrity is the most profitable equilibrium. You have to possess some organizing principle that's more foundational than the pursuit of profit to avoid it. There's ways to do that without getting totally wonky. You could organize as a Co-Op (so you operate group-selfishly instead of leader-selfishly), register as a benefit corporation, etc.
--
As for the other examples, what I like about tech is that anyone can join the industry. I think the way attorneys and doctors have set up their profession works for them but it really makes it hard for anyone who didn't choose the golden path to join.