This is an unpleasantly anti-pedestrian junction. There are signal crossings on two of the three major arms, but they are staggered (and "caged") two-stage crossings with long delays, meaning that getting from the Iffley Fields corner to Boundary Brook Rd can take four and a half minutes. So someone walking a child to Larkrise primary from Swinburne Rd (and doing this four times a day) could spend over an hour a week waiting at this junction! It is also pretty bad for cycling, most obviously in having no support for right turns from Iffley Rd and Donnington Bridge Rd.
Oxfordshire Liveable Streets has commissioned designs for this junction from CityInfinity.
Of the two options CityInfinity came up with, I think the first (shown above) is the most practical. (The second proposes a two-way west-side cycle track on Iffley Rd, but while that simplifies this junction it creates transition problems on either side.)
As with any junction design that properly supports walking and cycling, this would involve a reduction in peak motor traffic throughput. For most Oxford junctions that is problematic because of its effects on bus speeds, but here traffic volumes are low enough that I think this is acceptable, though it would need proper traffic modelling to confirm that. The lower traffic volumes here are a result of the congestion charge and East Oxford low traffic neighbourhoods, reducing peak flows along Iffley and Donnington Bridge Rds, and of the Larkrise School Streets scheme, reducing peak school-run flows into and out of Boundary Brook Rd.
This is also a "light-weight" redesign, with the carriageway reduced in width to provide space for cycling, so should be relatively cheap to implement (though that might just mean £500k instead of >£1m). Again, however, this would need to be confirmed by a full design considering topographic and geophysical constraints. This is not the highest priority junction in Oxford, either as a barrier to walking and cycling or as an injury hotspot, but it is one of the easier ones to fix.