I don't see this as a big problem.
Starlink is already reached the apex of its possible performance, what SpaceX will be trying to do no is rollout more satellite to bring more capacity and cover more oceans. A 100% increase in number of satellites won't equate to a 100% increase in service quality, it's now going to be more like 10-15%.
For Amazon Leo, true that they have less satellite, but we should note that majority of the cover Asia, Europe and North America and I believe that the majority of future satellite launches will be to improve coverage over this regions which is where about 90% of Delta's operations exist. Before the 2028 launch, I believe that Amazon will be able to have about 2,000+ satellites which is roughly about the same number starlink had at its introduction and still got slot of raves for its speed and latency.
The only thing Delta and Amazon needs to ensure is that they secure STC for most of the fleet before 2028 and work on making installation and rollout fast if not faster than what starlink currently achieves. Even if installation starts in 2029, and they can get the 4-days per plane installation timeline, I believe that they will be able to complete installation by 2031.
On the topic (Delta's reason was "silly"), taking the context and the competition into account, I'm not sure how a sensible person could disagree with Ben on this one.