Dual-funded channel opens successfully but peer then seems near-permanently offline/log shows "update_fee 253 unaffordable"

I set up a new node and tried to get some inbound liquidity using dual funding, but the channels I open appear to be useless and the peers show as offline almost constantly, although I don't think they really are. "lightning-cli invoice" nearly always gives a warning "Insufficient incoming capacity, once offline peers were excluded" and paying the generated invoice always fails.

I am running v0.10.2 on Ubuntu 20.04.4 on x86-64.

`getinfo` shows:
```
{
"id": "xxxx",
"alias": "xxxx",
"color": "xxxx",
"num_peers": 2,
"num_pending_channels": 0,
"num_active_channels": 2,
"num_inactive_channels": 0,
"address": [
{
"type": "torv3",
"address": "xxxx.onion",
"port": 9735
}
],
"binding": [
{
"type": "ipv4",
"address": "127.0.0.1",
"port": 9736
}
],
"version": "v0.10.2",
"blockheight": 733379,
"network": "bitcoin",
"msatoshi_fees_collected": 0,
"fees_collected_msat": "0msat",
"lightning-dir": "/home/lightning/.lightning/bitcoin"
}
```

My config file is:
```
proxy=127.0.0.1:9050
bind-addr=127.0.0.1:9736
addr=statictor:127.0.0.1:9051
always-use-proxy=true
bitcoin-rpcuser=xxxx
bitcoin-rpcpassword=xxxx
log-file=xxxx/lightning.log
experimental-dual-fund
plugin=xxxx/summary.py
```

I opened a new channel by executing:
```
lightning-cli connect 038fe1bd966b5cb0545963490c631eaa1924e2c4c0ea4e7dcb5d4582a1e7f2f1a5 xtdo5qvvfwcjaruj6z4acdcw4azagn6tdgac4ajnekjdn4ghr6qw2nqd.onion 9735
lightning-cli -k fundchannel announce=false id=038fe1bd966b5cb0545963490c631eaa1924e2c4c0ea4e7dcb5d4582a1e7f2f1a5 amount=10000sat request_amt=3000000sat compact_lease=029a00140000000001f4
```

This succeeded, but the peer almost always shows as offline when I use the summary plugin. Once in a while the summary plugin shows the peer is not offline, but this almost always reverts back to offline if I recheck the status straight after.

I thought this was a problem with this specific peer at first, but I have subsequently tried this again with a different peer and had the same behaviour - the connect succeeds and then the node appears to be nearly-permanently offline straight afterwards. Maybe I'm just selecting bad peers of course.

The log shows a lot of entries like this:
```
INFO xxxx-chan#3: Peer transient failure in CHANNELD_NORMAL: channeld WARNING: warning channel yyyy: update_fee 253 unaffordable
INFO xxxx-chan#3: Peer transient failure in CHANNELD_NORMAL: channeld WARNING: warning channel yyyy: update_fee 253 unaffordable
INFO xxxx-chan#3: Peer transient failure in CHANNELD_NORMAL: channeld WARNING: warning channel yyyy: update_fee 253 unaffordable
INFO xxxx-chan#3: Peer connection lost
INFO xxxx-chan#3: Peer transient failure in CHANNELD_NORMAL: channeld: Owning subdaemon channeld died (62208)
```

The node has on-chain funds (as shown under "utxo_amount" in the summary output) of over 100000 sats.

Am I doing something wrong or is this just bad luck with my peers?

The above is all factual, the following is just conjecture on my part in case it's helpful: I've been doing my best to figure this out and I'm suspecting that I've accidentally put too little liquidity in the channel on my side and my node can't therefore promise the peer that it will be able to pay for a cooperative close even at the minimum feerate of 253. If this is the case, it would be nice if the channel had failed to open rather than opening in a useless state.

If you need any more information please let me know. Thanks!