I would agree glycogen depletion is associated with fatigue - but I don't necessarily agree it is always associated with acidosis. For example - if you deplete glycogen content in muscle with prolonged, moderate-intensity exercise, you're unlikely to induce a significant strain on the pH of muscle.
Elite athletes will have glycogen values of ~700-800 mmol/kg dry weight muscle. Maximal intensity exercise (think 400-600m all out) has a rate of glycogen depletion of ~100 mmol/kg dry weight/minute, which drops to ~40 mmol/kg/minute for most mid-distance event durations. It has been suggested that you need to drop below 250-300 mmol/kg to impair function. This would suggest you are highly unlikely to deplete glycogen across the entire muscle in middle distance events lasting 100-780 seconds (1:40-13:00 mm:ss).
In contrast, it's entirely possible to deplete glycogen in an event like the marathon, despite lactate levels never exceeding 4 mmol. But yes you are correct - lactate production consumes a H+ ion, but this occurs regardless of where the glucose comes from (glycogen or blood glucose).
As for those speeds: 9 m/s is 540m/60s. That's ~44sec 400m pace.
·Hocker closed the lasted 300m in Paris in ~39.65, which is ~7.6m/s. El Guerrouj also closed the last 300m in ~39.77 when he ran 3:26.
·The fastest 100m in both the Wanamaker Mile (Cam Myers) and the Dr. Sanders 3000m (Grant Fisher) was slightly slower, at ~13.6 s (7.4 m/s).
·Rudisha's opening 200m during the 2012 Olympics was his fastest at ~23.4s, which is ~8.5 m/s.
So while top tier athletes they may potentially hit the speeds you've mentioned as an absolute peak speed, they are absolutely not holding 9-10 m/s for any significant portion of time even when running World/Olympic Records.