@ rcsa.
A second FDR is not as simple as it sounds.
Aside from the cost of a unit, ball park I doubt you'd get much change from $100k - remember that cost is the prime driver for not fitting them to light aircraft, you also have the additional wiring looms to run to wherever the second unit is fitted. This is extra dead weight that the vast majority of aircraft it is fitted to will have to carry around every time it leaves the ground and will never be needed, but which will have to be paid for by it's passengers. Admittedly, not a horrendous sum per ticket, but it's still an extra cost.
Then you have the issue of reconciling data between units - which one is the master, which is a slave & what happens WHEN there is a discrepancy between the data on each unit, what data do you believe.
Associated with this is the separation between the FDR's on the airframe, but fed by a common wiring loom. Even allowing for dual and triple redundancy, it isn't inconceivable that damage to an area of the aircraft could result in some sensor information only going to one recorder & different sensor data streams to the other. possibly the only way to avoid something like this happening would be to have a radio link between the units and record the other FDR's data on each FDR as well as its own, but where do you stop?
Your point about a floating FDR rising at 1000 ft/min is also not so simple. If we assume that in the worst case any debris has to go to the max depth expected on the route (seems to be about 4000m by general consent) then it will be subject to an external pressure not too far from 400 bar/atmospheres or 6000psi (all in very round figures). Most conventional forms of bouyancy simply can't cope with this without being crushed to the point of uselessness. To have bouyancy which is proof against this pressure you need to use fairly specialist material b ut the problem with it is that it is fairly dense.
To get a steady 1000ft/min (roughly 5.1 m/s) on a FDR box about 3 foot x 1 foot x 6 inches you would create a maximum drag force of about 8kN or about 800kg which would have to be balanced by the upthrust due to the density difference between the box and the surrounding sea water. Seawater is more dense than fresh which has a nominal density of 1000kg/cu. m, taking the box sizes above, even if it's totally empty, its volume is only .045 cu.m hence can only displace 45kg of fresh water or maybe up to abot 46 / 47 if you hit a really dense patch of sea.
The problem with the bouyancy materials which can take the pressure is that they are fairly dense, at somewhere between 800 and 950 kg/cu.m (generally solid epoxies loaded with different proportions of glass microbeads) - So unfortunately you wouldn't have a significant upthrust to get a floating FDR back to the surface PDQ without making the container unacceptably huge and heavy.
Sorry to be the bearer of bad news, but I'd say that if anything was to be adopted, a pack of dye under every 10th seat with different time release packaging would probably be favourite. Don't stop thinking though!