In my first attempts, YCast was served on the local IP of the pi, on port 8010. With this setup I had to wait about 2 minutes after each pi boot for YCast to run.
By staying on port 80 I see that YCast is executed instantly when PI starts up.
Out of curiosity, what is the point of using FastCGI?
Nevertheless I practiced doing the deployment by FastCGI. How do I check that I have succeeded?
YCast continues to work very well.
No error message by doing:
cat /var/log/lighttpd/error.log

My application:
#!/usr/bin/python3

Path to cloned YCast repository

PATH_YCAST = '/usr/local/lib/python3.7/dist-packages'

Path to my stations file

PATH_MY_STATIONS = '/var/www/ycast/stations.yml'

import sys
sys.path.insert(0, PATH_YCAST)

from flup.server.fcgi import WSGIServer
from ycast import server

if name == 'main':
server.check_my_stations_feature(PATH_MY_STATIONS)
WSGIServer(server.app).run()

My server:

Using FastCGI server for Ycast

$HTTP["host"] =~ ".(radiosetup|vtuner).com" {
fastcgi.server = (
"/" =>
(( "socket" => "/tmp/ycast-fcgi.sock",
"bin-path" => "/var/www/ycast/ycast.fcgi",
"check-local" => "disable",
"max-procs" => 1,
"fix-root-scriptname" => "enable",
))
)
}