Centralised discussion space for TfL plans and cycling in London

One thing to consider is the pressure to implement these schemes with as few interventions as possible, due to cost, time and the required experimental nature of them. Stuff like tidying up one-way systems can be done at a later date.

No. It needs to be done at the moment of inception. Otherwise, you'll wait for years and years. Returning streets to one-way in filtered cells is child's play.

Also, "keep cells as small as possible" is fine on paper, as long as you're happy to condemn one of the roads where you split a large cell to becoming a main road.

Yes, as I said, I fully recognise that Odessa Road doesn't have the right characteristics as a cell boundary street.

If you're worried about journeys within the cell being too long, that can surely be solved by making the sectors/quadrants/whatever* within the cell smaller.

I think what you may mean is 'make them as evenly-sized as possible'. This is what I was talking about when I said you have to filter in the centre of a cell. The larger the cell, the longer journeys will be within it even with optimised 'quadrant' sizes. There's no way around that, I'm afraid.

(* I've heard some people refer use "cell" to mean both the thing surrounded by main roads and the divisions within that, which makes discussing these things even more fun)

Ah yes, that can certainly cause confusion. 'Sub-cell' is a good word.

except for footway level entry tables

I put up with these every day as a pedestrian (Archway end of Holloway Road) and they're the absolute worst for ambiguous priority. Let's build a thing that encourages pedestrians to cross without stopping to look, but which drivers still have absolute legal priority over! Awful things.

Many are badly-constructed and at the end of unfiltered streets, yes. They're still better than 'continuous footways'. What makes the difference, as ever, is how drivers behave turning into them.