A phone speaker hides the low end and reveals the arrangement.
A phone speaker is a bad speaker and a useful one. It hides almost all of the low end and flattens the stereo image, which sounds like a disadvantage until you realise that arrangements are often easier to hear when the bass is gone.
What remains on a phone speaker is the midrange: the vocal, the guitars, the synth parts, the snare. Those are the parts that carry genre and era, and they are the parts a short clip is usually testing. A gated snare sounds like a gated snare on a phone. A doubled vocal sounds doubled. A cheap speaker is a midrange spotlight.
The trick is consistency. A phone speaker is fine as a daily device if it is your daily device, because you build a reference for how songs sound through it. Switching between a phone and studio headphones every day is what causes confusion, not the speaker itself.
There are rounds where a phone will fail you, and they are predictable: anything where the answer depends on a sub-bass line or a wide stereo effect. In those rounds the phone is not lying; it is simply not showing you the whole recording.
I use the phone for the daily round and speakers for learning sessions, and the split has worked for months. Try visit the site on your phone for a week and see whether the limitations show up where you expect, or somewhere more interesting. The phone is also the honest device for the way most people actually listen to music. If the puzzle's songs sound thin on a phone, that is how the songs sound to a lot of listeners, and understanding that is part of understanding the record. The game is about hearing, not about fidelity. The recording is the thing being tested, not the gear.