If you use a simple system in which you just mine for one type of block and the proportions of currencies/bonds in each block are constant, then you only need one difficulty. You can set this difficulty using the current algorithm. No modifications necessary. I could make a hard design which allows multiple mining choices and would need a new algorithm. I don't think benefits associated with a new algorithm are sufficient to justify added complexity.
In that case, you have to make sure to set the proportions and the maturity dates appropriately. It would be interesting to run some simulations and see if this is stable (i.e., reaches equilibrium).
Stability is not an issue. As in the current system, supply per unit time is fixed. Bonds at all maturity dates will sell for a non-zero price. The farther out the maturity date the lower the price. I'm 100% sure of this.
I can only guess about the actual prices. Assuming an interest rate of around 2.5% , then...
0.976 bitcoins for a bond with a face value of 1 bitcoin that matures in one year, maturing one-year in the future,
0.952 bitcoins for a bond with a face value of 1 bitcoins that matures in two years,
0.48 bitcoins for a bond with a face value of 1 bitcoin that matures in 30 years.
Approximately (ignoring risk aversion), the price of contingent claims will be the perceived probability of the difficulty level in the range specified by the bond multiplied by the price of the bond. For example, suppose that the market expects that probability that difficulty falls between 3 and 5 million one year from now is 0.5. Then (using the bond prices above), the price of the associated contingent claim will be
0.5 * 0.976 = 0.488.