First, let's go to the real issue: your proposal to limit the amount in an address does nothing to prevent whales from selling large quantities of bitcoins.
A person with 10000 BTC spread among 1000 addresses can still send all of those bitcoins to an exchange and sell them all at once.
And now why the implementation is problematic:
You wrote that they couldn't move their bitcoins because the hard fork would occur in the past. I'm saying that doing a hard fork in the past is not feasible.
The hard fork will happen in the future, but the wallet of the new fork will check all blocks after block 708077, if the tx outputs are valid. My opinion: it will work.
The hard fork happens at a specific block in the block chain regardless of when nodes start checking, so your hard fork would be at block 708077 regardless of when the change is actually implemented.
Here is an example of the problem:
A. I have an address with 100 BTC and I send 1 BTC from it to an exchange in block 708078.
B. Then, someone sends BTC from the exchange through Bitpay to CheapAir to buy airline tickets.
C. Then, Bitpay sells the BTC to someone.
D. Then, the person that bought the BTC sends it to Bitrefill to buy gift cards.
E. etc. ...
At a later date, you freeze my 100 BTC and invalidate transaction A in block 708078. As a result, transactions B, C, D, E, and so on will also be invalidated because they all are linked to transaction A. What happens to all those other people whose transactions are invalidated simply because my transaction was invalidated?
Also, how quickly do you think that the new rules can be implemented? Every day that it takes to switch to the new rules after block 708077 means another 144 previously valid blocks are later invalidated. Are miners going to mine any blocks after 708077 knowing that their block rewards will be invalidated?