But most of the negative comments are really baseless. The concept is odd and most people don't like stuff they're not used to.
I'm
very accustomed to radiant floor heating, and prefer to see the math. A radiant slab with embedded PEX isn't a way to quickly raise the surface temperature during a shower.
Even at the 99% design heat loads of a bathroom in central Florida the floor temps won't need to be very warm, making full radiant floor heating not particularly attractive, but sure you can do it, even using the water heater, though isolating the potable water from the heating water with a plate type heat exchanger would be preferable. That's not a particularly rare thing to retrofit in the northeast, where 99% outside design temps are in single digits F above/below 0F (compared to north of 40F for the Orlando area.) If one wanted to design a radiant floor heating system for the bathroom using only a small piece of slab/floor as the heat emitter so that the temps are warm enough to matter for bare feet that's fine, but don't use the hot water flow to the shower as the heating system water- it only adds complexity to the design.
And don't hack it- design it. Start with a Manual-J heat load calculation for the room, and adjust the radiating area & surface temperature needed to deliver the design-condition heat to the room with the amount of radiant slab you intend to have. Tweak the area such that even at design conditions (42F outdoors for Orlando) the surface temp of the patch heating the room isn't much more than 90F (or it will be the opposite of comfortable when it drops below the 99% outside design temperature), but not less than 80F (or the barefoot comfort benefit during average wintertime heat load conditions won't be enough to matter.)
Heating a small patch of floor intermittently & quickly for barefoot comfort while stepping out of the shower calls for an electric mesh under the tile solution, not a full-on slab heating solution (whether using a recirculation pump or not.) The time lag of the thermal mass of the concrete is real. That lag becomes considerably smaller if the heat is being applied 3/8" below the surface of the tile rather than 2-3" below. The efficiency hardly matters- in my area residential retail electricity is about twice that of Florida, but intermittent heating (or even the room heating) of small bathroom floors with resistance mesh heating usually pencils out favorably against hydronic solutions backed up by condensing gas boilers. The "payoff" of lower heating costs with a full room heating solution vs. just heating a patch of floor with electricity briefly when needed just isn't there when the space heating load for the room is so miniscule.