Like the use of D3 libraries, live variables and mapping functionality as in : https://youtu.be/uEmDwflQ3xE?t=513
My company is trying to break away from tableau 
how close can Observable get to a ‘radial search’ that is a visual that dynamically draws a radius on a map and filters the selection to that radius?
We can produce somewhat similar visuals in Python… could we embed some of that python code within a OHQ notebook?
You can certainly build this in Observable. You could use e.g. Leaflet (notebooks) or OpenLayers (notebooks) (or whatever other map rendering library/service) for drawing your map. The fields at the top you can make with Observable Inputs. Also the table at the bottom: Observable Inputs: Table.
Depending on the amount of data you have, you might want to use some kind of spatial data structure for implementing the radius filter (Vladimir Agafonkin of Mapbox (and original Leaflet author) has released some nice open-source libraries for this). But if you have fewer than say 1000 points you can just loop through them and measure the spherical distance from each to the target, and filter out ones with distance too far.
If you link to a notebook where you have imported your data and made your best effort at implementing the UI you want, someone might be able to help you with any parts you get stuck on.
For more general kinds of graphics, you might enjoy Observable’s recent Plot library. e.g. see Eric Lo’s notebooks about imitations of Power BI charts using Plot:
could we embed some of that python code within a OHQ notebook?
Conceivably you could run CPython through emscripten and get your Python code running in the browser (along the lines of this notebook), but I would instead recommend just reimplementing your code in Javascript. See:
