Running IPFS at scale is horrible. Try to download a few dozen TBs of small files. Its garbage collection is rubbish (ended up nuking the ZFS dataset every couple of days instead), it is very CPU and IOPS hungry, and it has bad network throttling support.
I would claim it has failed the test of time as it has very little adoption.
I would claim it has failed the test of time as it has very little adoption.