I would like to describe the reasoning by quoting Peter Ujfalusi
<peter.ujfalusi@ti.com> from his review of this patch series v1 [1]:

> When you bind a link you will use set_fmt for the two sides to see if
> they can agree, that both can support what has been asked.
>
> The pcm512x driver just saves the fmt and say back to that card:
> whatever, I'm fine with it. But runtime during hw_params it can fail due
> to unsupported bus format, which it actually acked to be ok.
>
> This is the difference.
>
> Sure, some device have constraint based on the fmt towards the hw_params
> and it is perfectly OK to do such a checks and rejections or build
> rules/constraints based on fmt, but failing hw_params just because
> set_fmt did not checked that the bus format is not even supported is not
> a nice thing to do.

[1] https://patchwork.kernel.org/project/alsa-devel/patch/
    20201109212133.25869-1-kmarinushkin@birdec.com/
Signed-off-by: Kirill Marinushkin <kmarinushkin@birdec.com>
Cc: Mark Brown <broonie@kernel.org>
Cc: Takashi Iwai <tiwai@suse.com>
Cc: Liam Girdwood <lgirdwood@gmail.com>
Cc: Matthias Reichl <hias@horus.com>
Cc: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Cc: Peter Ujfalusi <peter.ujfalusi@ti.com>
Cc: alsa-devel@alsa-project.org
Cc: linux-kernel@vger.kernel.org
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Link: https://lore.kernel.org/r/20201115122306.18164-4-kmarinushkin@birdec.comSigned-off-by: Mark Brown <broonie@kernel.org>