`Design_Matrix` fixes & enhancements

[ejolly]

We had also discussed how subclassing is handled #277 but this might be informed by the decisions we make in #437

Relevant issues:

• onsets_to_dm() weighting functionality #405
• consider making Design_Matrix more flexible with custom index #347
• Add a method to calculate optimal highpass filter cutoff in Design_Matrix #284