I hate to jump into this conversation, as I sort of lost interest in the physiological details some time ago. Before walking away from the subject, I read a lot from Stephen Seiler, including a time course model for improvements based on 3 waves of change, illustrated in one nice graph. For lack of better terms, and ease of understanding, hoping I don't misrepresent it, roughly they are:
1) VO2max -- this source of performance improvement comes rapidly, and after 1-2 years, most performance improvements have been realized, and start to plateau.
2) LT -- in the mid term, we get better at using a higher percentage of VO2max, sustaining it for longer. This takes longer, say 4-5 years, before most of the benefits from this source have been realized and start to plateau.
3) Economy -- ability to go faster with the same energy consumption. This comes from many years of practice, and eliminating wasted or conflicting movements, leveraging static properties (like tendon and muscle elasticity), and improved synchronization from the nervous system.
Right or wrong, I found it a simple and powerful model for athletes who are not Ex. Phys. Phds, to understand how training helps at several physiological levels, aerobic and non-aerobic.
Of course, if you detrain, you lose aerobic and economic adaptions, but they come back fairly quickly once training resumes. And maxing out VO2max doesn't mean you stop training it -- you still need "VO2max training" to maintain the high level.
Once you've attained a high level of "aerobic" fitness, the last long term phase is about improving from non-aerobic factors, like fine tuning the nervous system, and improving your form.
Elite athletes are already highly developed in all three. So while extra oxygen delivery to the muscles appears to help significantly in studies with amateurs, and slower athletes like a "clean" Lombard, the question becomes, how much improvement can elite athletes achieve, by increasing oxygen delivery in athletes who already have a high level?
Other topics to consider: if we get more oxygen to the muscles, producing more energy at a higher rate, heat is one by-product of this increased rate of energy production that needs to be managed/dissipated, and excess accumulation will eventually limit performance. I suppose this is more important for longer duration races. Another factor in long races like the marathon is how this impacts glycogen depletion, which will also limit performance in a big and sudden way. Another factor is the increased viscosity caused by red blood cells actually reducing blood flow, and the delivery of oxygen to the muscles.
Regarding steroids for women for strength events -- I think historical performances from the 70s and the 80s are hard to explain without such an effect.