How do I adjust the cooking time for two joints as opposed to one?

Cooking by time isn't always the most reliable way to judge done-ness. I would just use the times (as is), as a guidline, but insert a thermometer and aim for 135-140 F for medium rare.

The thermometer is the best guide. As there are many, many factors that can affect actual cooking time.

talon8 is absolutely right: the best way to go about this is to use a thermometer. However, it is still an interesting question how the expected cooking time varies with the weight of a roast.

It is such an interesting question that in 1961, SIAM Review published a scholarly article entitled "On Cooking a Roast". To be a bit more precise, the question that the authors answer is this: Suppose we have two roasts with exactly the same combination of identical tissues, and the same shape except that the one is a scaled (blown up) version of the other. Let's weigh both roasts and compute the ratio between the two weights; let's call that ratio r. Suppose furthermore that the roasts have the same starting temperature, and we will cook both until they have both reached some given (higher) temperature at their centre. What can we say about the times that both roasts need to cook for?

The answer is that under these assumptions, the bigger roast will need r^(2/3) times as long as the smaller one.

So, assuming each small roast (700 and 950 grams) has the same shape as the big one that the recipe writers used (let's say 1750 grams), for the 950 gram roast you'd expect to use (950 / 1750)^(2/3) = 0.665 or about two-thirds of the cooking time, and for the 700 gram one you'd expect to use (700 / 1750)^(2/3) = 0.543 or just over half of the cooking time. In particular, for the 950 gram roast, you should expect to roast it at 200C for 20 minutes, then turn down to 180 and roast for around 0.665 * 1750 / 450 * 12.5 = 32 minutes more. (1750 / 450 * 12.5 = 49 minutes is the average time they would recommend for the 1750 gram roast, and we're using that as the base for comparing our 950 gram roast against.) Then check with your thermometer!

Finally, I think for the resting time, you might as well use the original 30 minutes, because that's not (only) about getting the internal temperature to a certain level (although it does contribute to that - the heat will spread out through the meat), but also about things like the muscle fibers relaxing and the like, which are not covered by the article.

Your roasting time is going to be as if for a smaller roast. The time that it takes to roast something is based on the time it takes to get the internal temperature to "cooked" depending on how you like it. The bigger the roast, the longer it takes for the internal temperature to reach the correct temperature. It is based on the distance from the outside to the inside on a roast, so two 3 pound roasts will take a good bit less time than one 6 pound roast. By contrast, if you have a very flat thin 6 pound piece of meat, it will take less time to cook than the 3 pounders.

In your example, base your time on the smaller of the two roasts. THEN use an instant read thermometer (or get a probe thermometer that stays in the roast with the beeper outside the oven...you won't regret the purchase).

When your 700g roast is "done" pull it and cover it. When the larger roast is done (thermometer again) pull it and cover it. Or, alternatively, if you have people who like their roasts at different levels of doneness, finish the larger roast to the lower level of doneness, knowing the smaller roast will be done more.