Attempting to declare a host volume that is read-only with write permissions could use a better scheduler error

### Nomad version
v0.10.3

### Operating system and Environment details
Nomad vagrant box

### Issue

When attempting to declare a host volume for a task group as `read_only = false` when the volume is declared as `read_only = true` in the client configuration, the scheduler error could be more helpful.

The current output is:

```
+/- Job: "example"
+/- Task Group: "cache" (1 create/destroy update)
+/- Task: "redis" (forces in-place update)

Scheduler dry-run:
- WARNING: Failed to place all allocations.
Task Group "cache" (failed to place 1 allocation):
* Constraint "computed class ineligible" filtered 1 nodes
```

Something like
```
* Volume "readme" with source "nmd-r" and constraint "read_only = false" filtered 1 node
```

Or even something less specific but still volumes related
```
* Volume "readme" filtered 1 node
```

### Reproduction steps

**client.hcl**
```hcl
# ...snipped...
client {
enabled = true

host_volume "nmd-rw" {
path = "/home/vagrant/host-volumes/nmd/read-write"
read_only = false
}

host_volume "nmd-r" {
path = "/home/vagrant/host-volumes/nmd/read"
read_only = true
}
}
```

**example.nomad**
```hcl
job "example" {
datacenters = ["dc1"]

group "cache" {

volume "my-volume" {
type = "host"
read_only = false
source = "nmd-rw"
}

# This volumes causes the error, since no client has an nmd-r volume that
# allows writes.
volume "readme" {
type = "host"
read_only = false
source = "nmd-r"
}

task "redis" {
driver = "docker"

config {
image = "redis:3.2"

port_map {
db = 6379
}
}

volume_mount {
volume = "my-volume"
destination = "/foobar"
read_only = false
}

volume_mount {
volume = "readme"
destination = "/wat"
read_only = false
}

resources {
cpu = 500
memory = 256

network {
mbits = 10
port "db" {}
}
}
}
}
}
```

Reproduce the scheduler dry-run with client.hcl and example.nomad, then trace the host-volume placement path that reports "computed class ineligible." The work is complete when a read/write mismatch identifies the volume and source, with coverage confirming the improved scheduler error.

Written by the indexing model from the issue text.