Sitting on it wheels now.
A major issue arose. With the planned bolts tying the crossbeam to the main beam there was a lot of lateral movement in the crossbeam - a couple of inches fore and aft at the wheels. I had enough height difference to chop them down and still allow enough tilt. I chopped them down by 40mm from this...
to this.....
...and that has solved the issue. Despite being 8.8 high tensile bolts they simply had too much give in them and that extra length was a lever on the stainless main tube. Lesson for anyone following me - keep the distance between tubes as short as possible. The tubes now touch at a 32 degree lean angle which I expect to be enough. I can always get a bit more lean by winding the rod ends up though again that'll increase flex.
Speaking of flex that weld of the main pivot to the crossbeam will need some reinforcement. On a perfect road it'd be fine but a pothole at one end only will impart enough force on the crossbeam which will act as a lever on that weld to destroy it. Like Alan I'll add the same type of reinforcement there.
You can see the outer wheel leans more than the inner one.
This is a function of the lower arms having two inboard pivot points versus one on the top. The inner wheel is turning a tighter circle so extra lean is useful. The uprights were originally made to a length to support the longer pivot bolts. The main tube now rides higher and the control arms now have a higher mount in the middle than the outside. I was going to shorten the uprights to match but changed my mind.
With the original bolts the arms were parallel with the crossbeam and the inner mount set to match the outer at 22.5 degrees caster which is bang in the middle of the 0 to 45 degree max range for the steering. The control arms have to swing on the outer mount to steer at various levels whilst the inner mount is fixed. This means the effective length changes which in it's original long bolt form means that if you set camber to zero at 22.5 degrees steering you get increasing positive camber as you move from that central point in either direction. Now in short bolt form there's more movement away from the 22.5 setting as you take steering out but less movement away as you add more steering on. This is a better scenario I believe as the smaller geometry changes will occur at more extreme steering where I'll expect to appreciate the smaller movement more. I'll set it at 22.5 degrees with some negative camber with a view to only coming to zero camber as it's at zero caster.
Realistically I need the seat to do much more on it.