Handle OSM API Timeouts

[bhousel]

Followup on #1304
We already handle a bunch of different error conditions with the OSM API.
But when things get really bad, we don't even get a response from the capabilities or any of the other calls, just a connection timeout:

We should handle this situation too.

[bhousel]

This relates to #1451 too where I said:

While we're at it, the AbortSignal can now have a timeout. This is a thing that all modern browsers support now - details. The default fetch timeout in Chrome is 300 seconds, but we should probably drop most requests after like 5-10sec.

The timeout should be quicker.

Some data taken during today's DDoS storm:

2.4 minutes is kinda a long time to wait for a response 😆
After a few seconds we can cancel it and display the red bar of sadness.