1) I screwed up my earlier post and took the shortcut of referring to energy density and then providing numbers for specific energy ('gravimetric energy density' is a bullshit term). I work in a mass-focused world usually so I think in those terms, and for the most part I get away with it because for my chemistries they're pretty linear (near future chemistries not so much). I should have said: lithium-metal primary batteries (non-rechargeable) generally have much better specific energy than lithium-ion secondary batteries (rechargeable). A pedestrian CR123A is 275 Wh/kg, where the best lithium ion batteries in the world are sitting right at 260-265 Wh/kg.
CR123As are generally lithium manganese dioxide. I get to work with more dense chemistries a lot -- lithium thionyl chloride, or LTCs, are available up to 500 Wh/kg. These are much more volatile and are the reason I'm inherently nervous about primaries.
Lithium ion makes up for their low energy density (Wh/l) and specific ener by being rechargeable. Also, primary batteries generally have low power density (a limited rate of discharge in proportion to their capacity). Less than 1C (one times capacity, so a 1.5 Ah battery doing 1.5A) is the best you can get from most commercially available primaries, whereas I am literally ten meters away from a 25C continuous lithium ion cell (a 30Ah cell that will merrily dump 750A for as long as you want to to hold the wire).
2) If you try to recharge a lithium-metal primary battery, it might explode. That's the danger with primaries. There are protective devices built into some cells, and you might get away with it depending on the voltage when you start charging (I'll explain the mechanism if it's of interest to folks). I wouldn't worry about using primaries in a dual-capable flashlight as long as the charging was done outside of the light. If the light has a USB port and can function as a charger, I'd avoid putting primaries in it as a matter of principle.