RemixIT and Remixed2Remixed are domain adaptation-based speech enhancement (DASE) methods that use a teacher model trained under full supervision to generate pseudo-paired data by remixing the outputs of the teacher model. The student model for enhancing real-world recorded signals is trained using pseudo-paired data without the ground truth. Because noisy signals are recorded in natural environments, the dataset inevitably suffers from data imbalance in some acoustic properties, leading to subpar performance for the underrepresented data. The signal-to-noise ratio (SNR), inherently balanced in supervised learning, is a prime example. In this paper, we provide empirical evidence that the SNR of pseudo-data has a significant impact on model performance using the dataset of the CHiME-7 UDASE task, highlighting the importance of a balanced SNR in DASE. Furthermore, we propose adopting curriculum learning to encompass a broad range of SNRs to boost the performance of underrepresented data.