Transaction creation will also be affected when you're trying to send a large amount relative to the size of your inputs (sending 25k CDC with inputs of 50 CDC each will require 500 inputs). I can generally send between about 10-15k CDC at a time before my client starts complaining about the transaction failing, but it may vary for you.

To completely disable staking, set reservebalance to a value that's higher than your actual balance. You can set it to something impossible if you like, to make sure that value is never exceeded

in cloudcoin.conf:

reservebalance=99999999

On the commandline:

-reservebalance=99999999

If you mined that 300k CDC then you probably several thousand inputs, which the client has to hunt through each time it wants to stake. Gathering those inputs into something larger (by sending to a new address in, say, 5000 CDC chunks) should help. I've run into this problem myself with a wallet.dat that grew to 180MB after mining, staking, and transacting for over a year.

You can also use 'reservebalance' in the config file or commandline to force the client to leave some or all of your balance alone when searching for inputs to stake. It's the staking that causes the client to be unresponsive.

    "version" : "v1.3.2.0",
    "protocolversion" : 60006,
    "walletversion" : 60000,

Where are your stratum servers located?

Do we have the ability to set our worker's difficulty instead of relying on vardiff?

This is the POS part you are talking about I guess, it sleeps while there is no block found / mined..  This is normal behaviour afaik

nanosleep({0, 200000000}, NULL)         = 0
nanosleep({0, 200000000}, NULL)         = 0
nanosleep({0, 200000000}, NULL)         = 0
nanosleep({0, 200000000}, NULL)         = 0
nanosleep({0, 200000000}, NULL)         = 0
nanosleep({0, 200000000}, NULL)         = 0
nanosleep({0, 200000000}, NULL)         = 0

too bad... I was mining on it 24/7 ..... now there is no mine for it