Recently I tried (for 3 days only) orderless, consult, marginalia, vertico and embark.
I was amazed regarding the init time of this framework (its init time feels like zero). I was surprised to find, that I have more Ivy related packages installed, than I need for this vertico & Co framework. Also the documentation of vertico and Co is very comprehensible. Those different packages play really well together. And those packages provide quite some nice features, I haven't seen in Ivy (I haven't looked for new features in ivy for quite a while, because Ivy just works fine here).
Nevertheless, I'm back on ivy right now (whose configuration I lavish and care since about 5 years), because there are some minor things (with vertico and Co) I would have to take care of, before I feel at home.
Here I wrote down some of those minor things for anyone who is interested (This is not a request to "fix" those. Also, I did not file bug reports, because I haven't investigated enough. So please take all this with a grain of salt.).

1. regarding Tramps hostname completion, when set up like described at the bottom of the vertico readme.: Having multiple logins for a hostname results in multiple entries (like e.g. ssh:localhost) and marginalia only showing *ssh*. It would be useful to show the loginname also. And how about tramp chains (like ssh:user@remotehost|sudo:root@remotehost:)?

2. I missed partial completion insert for candidates, like used within bash.
   Example: 4 candidates foo-bar-1-x foo-bar-1-y foo-bar-2-y foo-buz-3 ; at the prompt ba has been written, pressing <tab> completes the prompt to foo-bar-.
   I know, it is not necessary for the completion to filter, but I'm only human and often like to read complete words instead of incomplete ones, especially, when I'm not sure which candidate I have to select ....

3. To be able increase the size of marginalia's area, would be really useful, on M-x, C-h f and C-h v.

4. When opening a bookmarked directory with consult-bookmark It would be nice to choose to get a dired buffer or but a find-file request inside this directory.

5. The # separator in consult-grep puzzled me somewhat, because I set up a different separator in orderless. But probably I misunderstood the use of it.

6. Getting a list out of consult-ripgrep with embarks list generating feature resulted in an endless loop without getting a result, but maybe this was just my configuration.

Edit: Point 6: it was not consult-ripgrep, but consult-line, where embark-collect-snapshot and embark-collect-live are stuck in an endless loop.
Edit2: Point 6 is an issue with my personal config. The functionality of Embark and Co works fine in a plain Emacs.