This doesn't seem to affect all Neon Testing systems, I have one older version where it works.

What seems different is that the failing systems have:

    $ systemctl --user status obex
    ○ obex.service - Bluetooth OBEX service
         Loaded: loaded (/usr/lib/systemd/user/obex.service; disabled; vendor preset: enabled)
         Active: inactive (dead)

and the working system has:

    $ systemctl --user status obex
    ● obex.service - Bluetooth OBEX service
         Loaded: loaded (/usr/lib/systemd/user/obex.service; disabled; vendor preset: enabled)
         Active: active (running) since Tue 2024-02-27 23:26:19 CET; 21s ago
       Main PID: 1154 (obexd)
          Tasks: 1 (limit: 9361)
         Memory: 2.0M
            CPU: 6ms
         CGroup: /user.slice/user-1000.slice/user@1000.service/app.slice/obex.service
                 └─1154 /usr/lib/bluetooth/obexd
         
    Feb 27 23:26:19 neon-testing systemd[774]: Starting Bluetooth OBEX service...
    Feb 27 23:26:19 neon-testing obexd[1154]: OBEX daemon 5.64
    Feb 27 23:26:19 neon-testing systemd[774]: Started Bluetooth OBEX service.
        
It's not evident what is forcing the service on or off. Just doing a:
    
    $ systemctl --user enable obex
    $ systemctl --user start obex

seems *not* to be enough. What seems new is that the desktop hangs until the service_start_timeout

Found this lead from https://bugs.archlinux.org/task/45816, dated 2015

Bug 481959 describes something similar

Also Bug 481805

See Bug 482082 which has the summary:

By default, the Breeze Splash screen is enabled. I noticed the system took a long time to log in after entering my password after upgrading to Neon / Plasma 6. I went to disable to splash screen in order to look at the text of what was being loaded and this time logging in to the desktop was almost instantaneous.

(In reply to tagwerk19 from comment #4)
> See Bug 482082 which has the summary:
> 
> By default, the Breeze Splash screen is enabled. I noticed the system took a
> long time to log in after entering my password after upgrading to Neon /
> Plasma 6. I went to disable to splash screen in order to look at the text of
> what was being loaded and this time logging in to the desktop was almost
> instantaneous.

A mentioned here: https://bugs.kde.org/show_bug.cgi?id=482082#c5

> There must be more than one issue in play here...
>   
> I've tried setting the splash screen to "none" (something that crashed systemsettings)
> in a test VM. Rebooting and logging on with Wayland and X11, I still get the org.bluez
> timeout and no difference to the logon delay.

(In reply to Arjen Hiemstra from comment https://bugs.kde.org/show_bug.cgi?id=482082#c7)
> Just to collect some extra data, can we the output of `systemd-analyze --user blame` for those that
> can reproduce this bug? While there is something going on with ksplash it'd be useful to exclude
> other things going wrong.
That was in the context of the splash screen mak a difference, seems not the case here, the result of "journalctl | grep blue":

    Mar 01 16:25:56 ... NetworkManager[383]: <info>  [1709306756.8001] Loaded device plugin: NMBluezManager (/usr/lib/x86_64-linux-gnu/NetworkManager/1.36.6/libnm-device-plugin-bluetooth.so)
    Mar 01 16:26:06 ... dbus-daemon[382]: [system] Activating via systemd: service name='org.bluez' unit='dbus-org.bluez.service' requested by ':1.36' (uid=1000 pid=727 comm="/usr/bin/pulseaudio --daemonize=no --log-target=jo" label="unconfined")
    Mar 01 16:26:31 ... dbus-daemon[382]: [system] Failed to activate service 'org.bluez': timed out (service_start_timeout=25000ms)
    Mar 01 16:26:31 ... dbus-daemon[382]: [system] Activating via systemd: service name='org.bluez' unit='dbus-org.bluez.service' requested by ':1.71' (uid=1000 pid=997 comm="/usr/lib/x86_64-linux-gnu/libexec/kdeconnectd " label="unconfined")
    Mar 01 16:26:56 ... dbus-daemon[382]: [system] Failed to activate service 'org.bluez': timed out (service_start_timeout=25000ms)
    Mar 01 16:26:56 ... kdeconnectd[997]: 2024-03-01T16:26:56 kdeconnect.core: No local bluetooth adapter found

and "systemd-analyze --user blame":

    1.526s xdg-desktop-portal.service
     961ms plasma-kcminit.service
     675ms xdg-desktop-portal-gtk.service
     319ms plasma-powerdevil.service
     188ms plasma-plasmashell.service
     184ms pulseaudio.service
     181ms plasma-xdg-desktop-portal-kde.service
     180ms plasma-polkit-agent.service
     144ms plasma-ksmserver.service
     134ms plasma-kded6.service
      75ms plasma-kactivitymanagerd.service
      49ms app-kaccess@autostart.service
      47ms app-org.kde.discover.notifier@autostart.service
      46ms app-lts_eol@autostart.service
      44ms plasma-kwin_wayland.service
      43ms app-org.kde.xwaylandvideobridge@autostart.service
      43ms plasma-restoresession.service
      38ms kde-baloo.service 
      33ms plasma-kglobalaccel.service
      27ms app-geoclue\x2ddemo\x2dagent@autostart.service
      19ms plasma-kcminit-phase1.service
      15ms session-migration.service
      11ms app-org.kde.kdeconnect.daemon@autostart.service
       9ms xdg-document-portal.service
       7ms dconf.service
       7ms xdg-desktop-portal-rewrite-launchers.service
       6ms at-spi-dbus-bus.service
       4ms app-snap\x2duserd\x2dautostart@autostart.service
       3ms xdg-permission-store.service
       3ms dbus.socket

The "blame" does not look particularly different between systems with the 25 sec bluez timeout and those without.

*** Bug 481805 has been marked as a duplicate of this bug. ***

*** Bug 482192 has been marked as a duplicate of this bug. ***

Git commit 83708cdb20448e908641e7a546e1a301532d9f96 by Harald Sitter.
Committed on 02/03/2024 at 12:14.
Pushed by sitter into branch 'Neon/release'.

temporarily disable the bluetooth link

it unreasonably blocks plasma startup
https://invent.kde.org/network/kdeconnect-kde/-/merge_requests/600#note_884500

fail the build to revisit this in April

M  +11   -0    debian/rules

https://invent.kde.org/neon/kde/kdeconnect/-/commit/83708cdb20448e908641e7a546e1a301532d9f96

Re-opening as this appears to be a general issue not specific to Neon.

A fix is in progress with https://invent.kde.org/network/kdeconnect-kde/-/merge_requests/643.

*** Bug 482538 has been marked as a duplicate of this bug. ***

*** Bug 482675 has been marked as a duplicate of this bug. ***

*** Bug 483247 has been marked as a duplicate of this bug. ***

Fixed by Simon Redman in https://invent.kde.org/network/kdeconnect-kde/-/commit/7f3287a71b7d1acd30724d05a0719312a5b8ae94 by disabling the KDE Connect Bluetooth backend for now. More targeted fixes are also in flight.

KDE Connect hasn't had a release, so at this point we're relying on distros to cherry-pick the fix, which has been recommended in https://mail.kde.org/pipermail/distributions/2024-March/001481.html. The flow of bug reports has tapered off, so it seems like that's been broadly happening.

(In reply to Nate Graham from comment #15)
> ... The flow of bug reports has tapered off ...
I think the earlier workrounds hadn't necessarily worked but seems fixed now.