Can MySQL use aggregate functions without GROUP BY?

I am a new data product, and I wrote a query today for a store that has not generated orders in the last 15 days.The SQL written by Data Development is 8 times faster than the results I wrote.
There is one place that I really don't understand, please ask for help humbly.

Written by me

SELECT s.id FROM `store`s JOIN ( SELECT `store_id` , MAX( `create_time`)t FROM `order` GROUP BY `store_id` )aON s.id = a.store_idWHERE s.channel = 2 ANDs.show = 1 AND DATEDIFF(now(), a.t ) > 15

Written by Data Development

SELECT idFROM `store`s where s.channel = 2 AND s.show = 1 AND DATEDIFF(now(), (SELECT MAX( `create_time`) FROM `order` o where o.store_id = s.id)) > 15

After reading the performance analysis of my SQL, it is because there are too many rows of order table data (about tens of millions), and the GROPU BY grouping is used to cause performance.
The development and writing avoided the use of GROPU BY, but my skills are relatively shallow, so I can't understand it:
In AND DATEDIFF(now(), (SELECT MAX(`create_time`) FROM `order` o where o.store_id = s.id)) > 15 In this query condition, the subquery statement outputs MAX time, but if group aggregation is not used, the output is notShould there be only one row of data?Why can I achieve my requirement of grouping by store_id?Although I probably guessed that the core lies in the following o.store_id = s.id, but I didn't understand it