No, I don't think the physics has changed since I posted this a year ago.

What you are proposing is diffusion which does happen, but on a much slower timescale than concerns us here.
Now first a caveat, because I assume a static tank, i.e. like a domestic hot water cylinder (HWC). As I said, if the contents are stirred, either by operating an immersion or by shaking the tank (moving boat) then stratification can be destroyed, mixing will happen, and a tankful of tepid water will result. If there is no stirring stratification holds up remarkably well. This can be easily verified using a thermal imager or by scanning up and down the HWC with an infra-red thermometer: the hot / cold boundary is very obvious.
To give a practical example: in our last house the HWC was a well insulated 210 litre unit with high and low level immersion heaters (IH). The high level IH was meant for "boost" purposes, i.e. heating a relatively small volume of water quickly (an IH can only heat water which is at or above its own level). Because there were normally only two of us living there, this facility was only used about 5 times in 20 years: initially to prove it worked, then for parties, etc. The low level IH was switched on for 4.5 hours in the early hours of the morning, this being sufficient to fully heat the whole tank from cold: normally, of course, there was residual hot water, so the thermostat switched off long before this.
My wife generally had a largeish bath in the morning, using about a third of the stored hot water. By late evening, with no additional energy input, water drawn off was only about 1 degC cooler than that drawn off in the morning, confirming both the quality of the HWC insulation and that stratification had been undisturbed.
HWC designers sometimes incorporate a baffle at the cold water entry point to direct the incoming stream slightly downwards. This helps to avoid mixing.