He is saying it is the not the same if you have 14 million in 1 tx opposed to having 14 million in 280 tx's (50k each), which is true. the following explanations are a bit simplified and not 100% accurate but still makes the point:

if you have 14 million in 1 tx then your chances of getting a selected stake are 1 in X. if you have 280, then it is 280 in X. to counteract not getting selected (since PoS is % based) if you only have 1 TX you will get more coins out of a single stake than you would if you have 280. X (as used before) is the number of all stake-able transactions on the network that are at least MinStakeAge old (2 hours for this coin).

the older the transaction gets the large a multiplier it will have from the coinage during payout hance why i said it a single tx would be paid more (older transactions should also have a higher chance of being selected for stake, but i haven't looked at this specific coin to see if that was changed). so if there are enough transactions ready to stake on the network (and coin age isn't a factor in selection) and you don't have enough in your wallet it is possible that you will never stake. if coinage is a factor in selection then it will just take a long time until your transaction gets old enough to be selected so it wont be that you will never stake, but that it will be very infrequent (like once a month or something like that, maybe longer since there is a hard coded block time for esp).


the possibility to never stake is much higher if a single individual or small group holds a large % of the stakeable transactions. which i guess i should elaborate on a bit.
I'm sure everyone is familiar with 51% PoW attack where you get 51% of the networks hash and then you control the coin because you will mine a large majority of the blocks.

PoS has a similar attack but it is a little different.
A 51% attack on a PoS coin is not based on if someone has 51% of the mining hash (since there isn't any) or total coins owned but rather 51% of the networks staking weight where stake weight traditionally is a combination of number of stake-able transactions and those transactions stake weight.
(this percentage actually turns out to be much lower, someone did the math a while back and its more like 30-35% or something, maybe less. id have to find the link on bct again, peercoin forum also has a bit of documentation on this).
how does high stake weight attack a network? So for example PoSwallet, they hold a large number of transactions and probably a decent portion of the network stake weight, it is possible that people give up staking on their own wallets since PoSwallet is getting all the stakes and then send their coins to PoSwallet in hopes that they now will get a portion of each stake (or whatever PoS wallets payout system is) and then soon PoSwallet is the only one staking and controls the network.

I am NOT saying PoSwallet would ever hijack a network they are staking for, I'm just using them as an example