Thanks to everyone for providing good arguements for improvement of the proposal. I have derived a second proposal and updated OP accordingly. If you have any counter-arguement to this proposal, feel free to put it here or in the comment section of the article -
http://upalc.com/maxblocksize.phpI don't thing it is a good idea to include TX fees in the calculation.
See this charts
{1},
{2} and
{3}Now consider and old miner consolidating his coins and a spammer attacking the network - both will cause increased volume of transactions (more or less in {1}) for a short period, but to succeed in the attack, the spammer (see 10 July and after) will include larger fees {3} and have less days destroyed {2}, while the old miner may 'donate' larger fees (as on 27 April) or use the fact that he may transfer them without fees (end of November), because they are old enough.
With your proposal of including the fees in the calculation the block size after 10 July will increase, thus helping the attacker even more, as it will keep increasing the block size (even now), just because others add more fees to mitigate the attack and prioritise their own transactions.