Feature request: Extend to time-varying coefficients (uber's orbit model)

[MarkusSagen]

In a blog post from PyMC marketing and Recast, there have been discussions regarding implementing time-varying coefficients to account for how seasonality effects marketing in a more deliberate way

• https://getrecast.com/seasonality/
• https://getrecast.com/uber-orbit/

It seems like Uber's Orbit time series model has a class for accomplishing this built in and PyMC has made an attempt to add this in their PyMC Marketing framework. https://app.reviewnb.com/pymc-labs/pymc-marketing/pull/250/

Would it be possible to add or implement something similar in lightweightMMM as it is right now? If so, is that something I could help with or get some starters on how to do?