Abridged) We model the X-ray properties of millisecond pulsars (MSPs) by considering hot spot emission from a weakly magnetized rotating neutron star (NS) covered by an optically-thick hydrogen atmosphere. We investigate the limitations of using the thermal X-ray pulse profiles of MSPs to constrain the mass-to-radius ($M/R$) ratio of the underlying NS. The accuracy is strongly dependent on the viewing angle and magnetic inclination. For certain systems, the accuracy is ultimately limited only by photon statistics implying that future X-ray observatories could, in principle, achieve constraints on $M/R$ and hence the NS equation of state to better than $∼$5%. We demonstrate that valuable information regarding the basic properties of the NS can be extracted even from X-ray data of fairly limited photon statistics through modeling of archival spectroscopic and timing observations of the nearby isolated PSRs J0030+0451 and J2124--3358. The X-ray emission from these pulsars is consistent with the presence of a hydrogen atmosphere and a dipolar magnetic field configuration, in agreement with previous findings for PSR J0437--4715. For both MSPs, the favorable geometry allows us to place interesting limits on the allowed $M/R$ of NSs. Assuming 1.4 M$_{⊙}$, the stellar radius is constrained to be $R > 9.4$ km and $R > 7.8$ km (68% confidence) for PSRs J0030+0451 and J2124--3358, respectively. We explore the prospects of using future observatories such as \textit{Constellation-X} and \textit{XEUS} to conduct blind X-ray timing searches for MSPs not detectable at radio wavelengths due to unfavorable viewing geometry. Using the observational constraints on the pulsar obliquities we are also able to place strong constraints on the magnetic field evolution model proposed by Ruderman. Comment: 9 pages, 7 figures, published in the Astrophysical Journal (Volume 689, Issue 1, pp. 407-415)