Coment on Figure 10

[neko-suki]

First of all, thank you for writing great article to explain HoL Blocking issue.

It is very clear and really understandable for me from the beginning to the first half part of HOL blocking in HTTP/3 over QUIC

However, it took more time to understand the middle of Figure 10.

You described we profit from QUIC's HOL blocking removal.

However, the sentence before it describes However, as discussed before, the browser probably won't benefit from that all too much if it's a JS or CSS file, as there is more purple data coming in the second flight..

So, I had confused what is the profit from QUIC's HOL blocking removal since the browser won't benefit from that.

In addition, I can't imagine how the middle approach is better than the top approach when packet is lost in this example. Intuitively and generally, I think it's true. So, further explanation may help reader to understand how it is good when packet is lost.

I thought how to improve it for few hours, but I have no idea for the moment.