Problem: Pruned node syncs really slow and uses more space

```
# From the doc
Pruned node (setting pruning=everything)
Storage: ~25G*
RAM: 4 GB (LevelDB) or 64G RAM (RocksDB)***
CPU: 4-core
```

```
# config.toml
db_backend = "goleveldb"

# app.toml
pruning="everything"

# mem
free -h
total used free shared buff/cache available
Mem: 15Gi 9.5Gi 185Mi 0.0Ki 5.6Gi 5.6Gi
Swap: 0B 0B 0B

# cpu and iostat are seems ok

# disk
/dev/nvme0n1 1007G 468G 488G 49% /data
```

I have been using cronosd (v0.6.10) to syncs from 1 to 2,693,800 for over one week. It seems too slow for syncing. Is it normal for a Pruned node? It has used 468G disk space which is more than 25G. Am I doing something wrong? Thx.

Start by reproducing the report with cronosd v0.6.10 using db_backend="goleveldb" and pruning="everything", syncing from height 1 to 2,693,800. Compare sync time and disk usage with the documented ~25G expectation, then determine whether the behavior is normal for a pruned node or identify the relevant pruning or database issue.

Written by the indexing model from the issue text.