The only reference I have ever come across concerning the timing of our valves is in the article "Super Speed and How to Obtain It" (see Motor Cycling of 17.8.1921, or this website).
Although there is no facility to adjust the timing to finer than the nearest camshaft-pinion tooth (about 8 1/2 degrees) - in common with 99.99% of all engines - it surely remains of interest to know what is actually going on.
With the engine/gearbox unit in the frame, measuring the valve timing is difficult, because the timing-chest cover needs to be in place to support the front end of the camshaft, and this prevents access to the nose of the crankshaft. About the only accessible degree-plate drive-point is the nose of the magneto, after removal of the large domed plug in the cover.
A degree-plate of about 5" diameter should be attached to the magneto via a distance-sleeve of about 1 1/4" length, and a central M6 bolt (what!) of about 1 1/2" length. A pointer can be held in place by one of the lower cover screws.
Because the magneto (and camshaft) rotate at half crankshaft speed, it will simplify life if the degree-plate is marked off 0* to 720*, instead of the usual 0* to 360*. (Note that this will cover two TDCs and two BDCs).
About the only accurate way to locate BDC is with a Piston Stop. This is a finger or prong of some sort that projects into the cylinder via the sparking-plug hole (and is ideally firmly located to it). Rotating the crankshaft forwards and backwards until the piston meets the Stop (preferably at about mid-stroke) gives two degree-plate readings. Half way between these is BDC. After that, the Stop should be removed, and valve opening and closing positions can be recorded, and set onto a Valve Timing Diagram.
At this stage, you are very possibly in for a shock...your Diagram won't look much like the reference Super Speed picture! However, if your model is performing to your satisfaction ("if it ain't broke"), don't (try to) fix it!
Thoughts on Sopwith Valve Timing
The "Super Speed" article lays great importance on getting the valve timings for both cylinders identical: but surely it would be better to get maximum power from each cylinder, irrespective of timings!
Anyway, getting the timings identical isn't everything. What goes on between valve openings and closings is quite complex. The rates of opening and closing, and the maximum valve lift and its duration ("dwell") all have their effects on performance.
The Sopwith timing-chest is equipped with four (identical) pivoted lobed-finger type cam followers. This is fine, except that the fingers for one cylinder trail the cams, and for the other cylinder, they lead. This means that the valve motions for left- and right-hand cylinders are different.
I don't think that it is possible to devise a single follower-lobe shape that would confer identical motion to each of the inlet- and exhaust-valve pairs. In theory, (different) lobe profiles could be developed to achieve this, but it wouldn't be a trivial task!
So, rather as per the last sentence above, if you are satisfied with your model's performance, feel comfortable about letting well alone!