The ballast compatibility is published on the lamps datasheet. If the lamp is rated as compatible with a given ballast type, there should not be any issue.
Whether a penning mixture, or any other HPS lamp is compatible with a constant current CWA ballast does not have to always be related to the lamp chemistry itself, but also to the lamp construction, especially its thermal management.
The fact is the constant current ballasts combined with a saturated vapor lamp in its basic form does lead to thermal instability (hither temperature -> higher pressure -> higher voltage drop -> higher power -> higher temperature positive feedback), but manufacturers came with design tricks to eliminate that sufficiently enough so the lamp becomes stable even on a CWA ballast. One example is to put the amalgam reservoir (aka from the vapor pressure perspective the "cold spot" whose temperature dictates the operating pressure) further away from the active arc, so suppressing the thermal feedback, but that is not the only one.
So the question is, whether any such thermal feedback suppression is used in the lamp design and whether it is effective enough to make the lamp really stable so compatible with the constant current drive of a CWA ballast.
From marketing perspective, if a big portion (maybe even a majority) of MV ballasts on the target market are the constant current drive CWA type, I would be very surprised any lamp maker would make a lamp advertised as a "MV replacement" without making sure its design is compatible with the CWA ballast.
Of course very different situation would be when you import the retrofit lamp from a "230V" area where series chokes are 99% of the MV ballasts in use, so where the saturated vapor lamp is stable by itself, so without any such thermal feedback suppression feature. Then you have to really make sure the ballast has the same or similar characteristics as the original series choke (so e.g. the HX transformer or so).
you say that, but the first generation of HPS retrofit lamps sold in the US where indeed incompatible with CWA ballasts, it was one of the big limiting factors when it came to the wide spread adoption
if you look up the GE E-Z-Lux and Sylvania Unalux, you will see how their entries in the catalogs and such like are plastered with warnings about being suitable for operation only on HX-Autotransformer or Choke type ballasts
the other main problem is besides general instabilities as your describe is that at the end of life the voltage of a HPS lamp rises and on a constant current ballast like a CWA ballast, this can lead to dangerously high power levels
it was not until unsaturated HPS lamps where devloped did CWA compatible HPS retrofit lamps become possible
dedicated CWA HPS ballasts I understand work around the issue by having a very low OCV, so the lamp arc voltage cannot rise that much before it exceeds what the OCV can support and the end of life lamp will thus safely cycle
