Staff model HMOs do pretty well in terms of streamlining the experience and balancing the provider-patient aspect. The downside is that there aren't very many of them, and there end up being limitations to choice - the patient can only see doctors that belong to that HMO system, and those doctors serve nobody except the patients of that HMO product (with exceptions for ER).
Contrast that with a contracted HMO, which is an altogether different beast. Staff model doctors are salaried, and the parent insurer takes the claim risk -
if there end up being more sick people than expected, it gets taken care of through parent insurer capital and reserves, the doctors aren't at risk In a contract HMO, however, the doctors aren't salaried - they're capitated. They get paid a fixed cost per enrolled patient per month. The capitation is based on an average expected utilization level. So if the doctor ends up with sicker patients who need more care, they end up footing that bill, and the cost of that care comes out of their profit margin.
Staff model HMOs have many fewer perverse incentives, but they're slow to grow/shrink as the patient population shifts. They're also geographically bound with very limited options for expansion.
Capitated contract HMOs are subject to a lot of perverse incentives, and patient experience is often much worse. They also end up being very risky enterprises, because the patient base is small so pooling of risk doesn't work so well.