What I have found in writing my own utilities is that failure messages are the most helpful when they indicate:

1. What is the error/status code response
2. What was the command/invocation that failed
3. What is a common cause or fix for that situation

In this case: how can I on the command line test the connection tracker is having refused?

tracker-control-general.c
                manager = tracker_miner_manager_new_full (TRUE, &error);
                if (!manager) {
                        g_printerr (_("Could not start miners, manager could not be created, %s"),
+                                   error ? error->message : _("No error given"));

This command works so dbus is available to the normal user
gnome-open /home/

This call fails:
 manager = g_initable_new (TRACKER_TYPE_MINER_MANAGER,
 NULL,
 &inner_error,
 "auto-start", auto_start,
 NULL);

with the gio error enum
G_IO_ERROR_CONNECTION_REFUSED: Connection refused.

So whatever is on dbus instantiating TRACKER_TYPE_MINER_MANAGER objects is no longer with us

logout/login does not fix this problem.

There seem to be something broken with dbus, in particular, the following command fail:
netstat --protocol=unix --listening --numeric --program | grep $DBUS_SESSION_BUS_ADDRESS

so programs connect using
DBUS_SESSION_BUS_ADDRESS=unix:abstract=/tmp/dbus-7s51hr6DcF,guid=ee01c17b14dbcc77065ef34657a79dfd
- this has to be checked in a terminal window under the desktop manager

but dbus.service does not listen to that socket (it does on a functional system)

Ubuntu also prevents restart of dbus
systemctl restart dbus.service 
Failed to restart dbus.service: Operation refused, unit dbus.service may be requested by dependency only.                                          
See system logs and 'systemctl status dbus.service' for details.                

and reload does not help

reboot it is

on login:
tracker prcesses are up, but no search data i available and no mining activity is taking place

things like
tracker daemon --watch
fails with the refused error message

reboot the kernel does bring the system back to normal operation.

Here is the proper command to verify that the dbus socket is available
sudo netstat --protocol=unix --listening --numeric --program | grep $(sed --silent 's/[^/]*\([^,]*\).*/\1/p' <<<$DBUS_SESSION_BUS_ADDRESS)
unix  2      [ ACC ]     STREAM     LISTENING     29658    4900/dbus-daemon

Could anybody be helpful and provide the proper troubleshooting steps for the situation:

How do you determine if the TRACKER_TYPE_MINER_MANAGER factory is dead?

How do you bring it back to life?

How do you restore dbus with something less than a full reboot?

Have this again.

I believe when the TRACKER_TYPE_MINER_MANAGER instantiator dies,

that when Tracker stops indexing
and Can no longer be restarted

Only known fix is power off

nothing is listening to $DBUS_SESSION_BUS_ADDRESS

So this is caused by dbus being broken.

SYMPTOM:
Desktop Search/tracker stops finding files and cannot be restarted
- casued by incorrect value in $DBUS_SESSION_BUS_ADDRESS

DIAGNOSE
qdbus
Could not connect to D-Bus server: org.freedesktop.DBus.Error.NoServer: Failed to connect to socket /tmp/dbus-apJquVdO08: Connection refused
- should list 100 or so apps that are listening to the user's dbus

The X session has an incorrect value in $DBUS_SESSION_BUS_ADDRESS

GETAROUND for Ubuntu 16.04 dbus 1.10.6
1. make sure dbus works
ps -fC dbus-daemon | cat
UID        PID  PPID  C STIME TTY          TIME CMD
message+  3174     1  0 Aug26 ?        00:01:24 /usr/bin/dbus-daemon --system --address=systemd: --nofork --nopidfile --systemd-activation
foxyboy  28549     1  0 Aug26 ?        00:01:35 /usr/bin/dbus-daemon --fork --print-pid 5 --print-address 7 --session
foxyboy  28592 28587  0 Aug26 ?        00:00:04 /usr/bin/dbus-daemon --config-file=/etc/at-spi2/accessibility.conf --nofork --print-address 3
- there should be one with 7 in the arguments
- if that does not exist, it hould be recreated by logout/login
- in this case process id is 28549

2. find the socket name
sudo lsof -p 28549 | grep /tmp/dbus-
- prints over 100 lines all with the same socket
dbus-daem 28549 foxyboy  131u     unix 0xffff8802dd100f00      0t0 12528592 @/tmp/dbus-3bFHfv2e6Q type=STREAM
- so socket is /tmp/dbus-3bFHfv2e6Q

3. update DBUS_SESSION_BUS_ADDRESS
export DBUS_SESSION_BUS_ADDRESS=unix:abstract=/tmp/dbus-3bFHfv2e6Q

4. should now work
qdbus
org.freedesktop.DBus
- lists 300 lines of apps listening to sbud

ps -fC dbus-daemon | cat
UID        PID  PPID  C STIME TTY          TIME CMD
message+  3174     1  0 Aug26 ?        00:01:24 /usr/bin/dbus-daemon --system --address=systemd: --nofork --nopidfile --systemd-activation
foxyboy  28549     1  0 Aug26 ?        00:01:35 /usr/bin/dbus-daemon --fork --print-pid 5 --print-address 7 --session
foxyboy  28592 28587  0 Aug26 ?        00:00:04 /usr/bin/dbus-daemon --config-file=/etc/at-spi2/accessibility.conf --nofork --print-address 3

for older versions, the socket is stored in the only file in directory ~/.dbus/session-bus

for newer versions, the socket is stored in /run/user/$(id --user)/dbus-session

dpkg --status dbus | egrep "^(V|Pa)" && lsb_release --description --codename
Package: dbus
Version: 1.10.6-1ubuntu3
Description:    Ubuntu 16.04.1 LTS
Codename:       xenial

Here is the one-liner that fixes your current shell:

_PIDS=($(ps -C dbus-daemon --format pid=,ppid=,user=,args= | egrep "^ *[0-9]* *1 *$(whoami) " | sed --silent 's/^ *\([0-9]*\).*/\1/p')) && if [ ${#_PIDS[@]} -ne 1 ]; then echo >&2 "failure: found ${#_PIDS[@]} dbus pids"; false; fi && _PID=${_PIDS[0]} && _FILE=$(lsof -p $_PID 2>/dev/null | sed --silent 's/.*@\([^ ]*\) .*/\1/p' | head -1) && if [ ! $_FILE ]; then echo >&2 "Failed to find dbus-socket"; false; fi && _CMD=(export DBUS_SESSION_BUS_ADDRESS=unix:abstract=$_FILE) && echo "${_CMD[*]}" && "${_CMD[@]}"

on success it echoes like:
export DBUS_SESSION_BUS_ADDRESS=unix:abstract=/tmp/dbus-ZMtofqapHn

and then  you can do:
tracker daemon --start
Starting miners…
  ✓ Userguides
  ✓ File System
  ✓ Applications
  ✓ Extractor

fixed 3.22.2