when we talk about scale, we specifically mean that lookup continues to scale past 20k notes. by scale, i'm referring to both machine and human performance. the reason is the hierarichal structure of your notes: this partitions the number of items that both the human and the computer have to filter to get to a specific note. this is actually more important on the human side (a computer can hold 20k entities easily, a human will need to chunk it into something smaller)