Shortly after the results were announced, various papers appeared which attempted to explain the observed lack of a second peak. These take advantage of the many free parameters which are available in models of the microwave background. One solution is to increase the baryon content rather than reduce the CDM content. In order to retain CDM one significantly violates either big bang nucleosynthesis constraints (Tegmark & Zaldarriaga 2000TZ) or cluster baryon fractions, or both. These were critical pieces of evidence which led to ; it is not a trivial matter to dispose of them in order to force the new data into compliance with the model du jour.
Another solution is to somehow erase the peaks subsequent to the first. This can happen if the microwave background photons encounter a significant optical depth, which requires substantial reionization at quite early times (Miller 2000Cole; Peebles, Seager, & Hu 2000PSH). How this could come about is unclear. There may also be decoherence of the ideal signal (White, Scott, & Pierpaol 2000WSP), in which case the microwave background will retain little information of interest beyond the position of the first peak.
These effects were not expected, and it is not necessary to invoke any of them if CDM does not exist. The small observed amplitude of the second peak is natural and expected. Nevertheless, any of these effects could occur. The physics is the same in either case -- the only difference is the presence or absence of CDM. It is much easier to explain the low observed amplitude of the second peak without CDM. That does not mean optical depth or decoherence or some other mundane effect need not matter in the purely baryonic case.
At present, a simple universe devoid of CDM suffices to explain the data. If the universe remains simple, the pure baryon case continues to make clear predictions. As data accumulate, the second peak should become clear. It is only marginally suggested by the data so far, but it should resolve into the shape predicted by the models in Fig. 1(b). The amplitude of this second peak will be smaller than the expectations of models. Beyond this, the power spectrum should continue to roll off to smaller angular scales so that the third peak has a lower absolute amplitude than the second.