David, I don't know if any of the following is going to be of any practical use to you but here goes.

Tower is wrong to claim that the government has mandated that the land in the red zone cannot be built on. If that was the case you would have received a "Notice of intention to take land" issued under s54 of CERA 2011 and you would be entitled to be compensated at market value under subpart 5 of CERA.

What the government has done is taken note of the fact that the deformation of the bedrock in the Feb 22 earthquake has reduced the ground levels alongside the Avon significantly so that houses consented since 2004 will not comply with the building code unless there foundations are raised and that they will not be insurable unless that happens (see Part 2 of the City Plan [online] for details of this being measured relative to the CCC Datum Point inthe estuary). The very arbitrary seeming way that houses in some streets have been allocated as red and green argues very strongly against the social dislocation and service disruption arguments being the main reasons for the government offering to buy out home owners in the red zone but not their neighbours in the green zone. Most importantly, note that these are offers of financial assistance rather than compulsory purchase orders.

While it does not address the ethics of Tower taking money from you for full replacement cover and then refusing to deliver it may well be that your best course of action will be to take their money and either repair the house or bank the cash and then sit tight until CERA is compelled to comulsorily take your property at fair market value.

Plus you'll need the council's Waterways, Wetlands and Drainage Guide to give you the baseline figures for the RL, recorded tide and rainfall levels, etc.
http://resources.ccc.govt.nz/files/Appendices2.pdf

Regarding the comparison with the 3m tsunami, I do have the original GIS files from the tsunami evacuation study because that study is complete and in the public domain. None of the information gathered on the lateral displacement of the riverbanks is allowed to be made public but you can replicate this for yourself with a tape measure if you measure the displacement at the bridge abutments and the width of cracks on the riverbanks and along the neighbouring roads. The main point of doing this is to measure how far the banks have intruded into the waterway as this allows you to calculate the amount of riverbed heave without having to get wet. Once you know how much the profile has changed you can calculate the reduction in river capacity when springtides in the estuary essentially act as a damn and cause the rainwater runoff to backup along the river.
If you can measure how far the cracking extends from the riverbed and the total width of the cracks then you can calculate the average amount of subsidence in that area due to lateral spreading. There are currently no official figures for the amount subsidence due to the removal of liquifaction silt but anybody who helped remove it can give a reasonable estimate, probably varying dramatically from property to property but generally 1-15 cm was my observation. A similar amount needs to be provided for the effect of compaction.

When all this is added together it adds to a flood risk (defined by the RL) covering the same surface area as a 3m tsunami (but without the damage caused by the speed of inundation in a tsunami).

There has been both uplift along the Heathcote and subsidence along the lower Avon as result of the Feb 22 quake which will tend to focus the springtide effect into the Avon but the size of that effect is to difficult to estimate.
http://www.starcanterbury.co.nz/local/news/map-shows-quakes-fury/3952900/

The cost comparison only came close to parity with the cost of relocation because the uplift and subsidence has a big impact on gravity sewerage system and assuming that the rebuilt local roads will need to have gravel bases normally only used on main roads due to the next Alpine Fault quake currently having a 80% probability of occurring within the normal design life of local roads. It is probably over 95% for the design life of underground infrastructure.

But note that the rebuild cost is the same as the buy-out cost only because the infrastructure cost is assumed to have already been included in the subdivision section prices. If that's not the case then, along with the social costs of forced relocation, the relocation will be more expensive than the rehabilition option. But I think you'll find that social costs are something that governments are not good at with infrastructure decision making.

I think that there are two important points here.

The general level of the riverside land is lower than it was before the earthquakes and the riverbed is higher than it was. That means that the area of land likely to be flooded in the event of a springtide coinciding with a heavy rainfall event is close to what would have been expected from a 3 metre tsunami prior to the earthquake changes. In that event the entire expressway from Bromley to Shirley Golf Course would be underwater along with half of Avonside, Dallington, Avondale, Aranui and South Brighton.

Once the cost of flood protection/mitigation works are added to the land remediation costs it may be cheaper top build new roads and pipes for new suburbs on more stable soils than to rebuild the existing pipes and roads with a deep enough basecourse to resist a couple of minutes of shaking from an Alpine Fault earthquake.

IMHO the costs probably will turn out to be similar so the decision really comes down to the social and community costs of the two alternatives.