sorry for taking time to answer – and it is going to be incomplete anyway (I had other things to attend to which required more attention yesterday).

You ask about the sensitivities. It is a rather technical question if interpreted the way I do. Sensitivity is not just resolution: it has to do with efficiency, noise, acceptance. A number of issues. In any case:
- the tracker of CMS is made of silicon pixels (the three innermost layers, closer to the beam) and strips (the rest). It is a wonderful instrument from almost any point of view. Tracks will be reconstructed with high efficiency and small noise even in the high-luminosity conditions. The rationale of building a tracker with silicon rests in several advantages: speed of response, insensitivity to background levels, redundancy; but the final goal is a high accuracy in the measurement of impact parameter, that is the distance between the primary interaction vertex and the track trajectory. An impact parameter significantly different from zero tags a track coming from a secondary vertex, produced by a long-lived particle such as a tau lepton or a b-meson. On impact parameter the tracker sensitivity is of the order of a few tens of microns, depending on track geometry and momentum – quite appropriate to the task.
- the e.m. calorimeter has excellent energy and spatial resolution, to reconstruct the decay over huge QCD backgrounds. I do not remember actual numbers – I can dig them out if you insist – but it is of the order of 1% for a 100 GeV photon.
- the hadronic calorimeter is less good: lots of money went into the former two apparata, and the muon system cannot certainly be sacrificed… Jets are measured with a resolution which is worse than that of other recent HEP experiments, I think of the order of 130%/sqrt(E) – which means 13 GeV resolution for a 100 GeV jet. Again, I am quoting off the top of my head; I will come back and refine these numbers if they turn out to be unacceptably off. Jets, however, can be measured also with the tracker and e.m. calorimeter to some extent (there are methods to do that), and a big effort in CDF is trying to tackle that issue. The most pressing issue is the resolution on missing transverse energy, which is measured in the hadronic calorimeter and is critical to detect neutrinos or other non-interacting particles. This, in my opinion, is the weak spot of CMS.
- the muon system is redundant and composed of different systems. The spatial resolution is not an issue, because it exceeds the typical spread in position which results from multiple scattering in the large amount of material the muons have to traverse before reaching the chambers. Redundancy and fast self-triggering are the strong points of the system.

As for the cost of a crystal, I have no idea, but it is high.We are talking of small change if compared to the typical budget of NASA missions, but it did weigh in the total CMS budget.

Cheers,
T.