Partially yes, if you have short reporting period, the link adaptation can adapt faster. However practically for very fast changing conditions, the “transmission -> feedback -> next transmission” loop may be too slow and it’s better to average the feedback and use lower order MCS.
The second thing is, that if there’s an application which requires relatively low throughput to fulfill QoS requirements, using fast adaptation rate (and thus low periodicity for feedback) will require high throughput for the feedback itself, and thus is not very efficient, in that case sometimes it’s also better e.g. to use larger period for feedback to have relatively good overhead/throughput ratio.
