but i checked and it's still the number in ontap 9 documentation
There's a few internals that i can think of that are depended on file count.
1. Quota. re-initialize quota on volumes with large numbers of file can take hours.
2. Snap vault. transer time can take longer.
3. deswizzling. when large snapshot deleted (happens after snapmirror transfer for example) the disswizler have to run on all the inodes. if you have frequent replication. the scan may never end as it re-starts after each snapshot deletion.