That's cool.  I was thinking of setting up a pool myself since I already burned the SEM on a delegate, but I haven't had a chance yet to writeup the software to split up the rewards and don't want people voting for me until I've had a chance to write the code and test it.

In the mean time, it's fun to do what-if scenarios.

Please let me know if my math is correct on this:

With 30 second blocks, and 1440 minutes in a day, there should be 2880 blocks minted per day in the network.  With 64 validators in testnet, that means each validator gets roughly 45 SEM (assuming testnet 1 SEM block reward).

Your semuxpool_com has about 62,224 votes now.  The semux.info pool has about 347,239 votes (including 1400 of mine) split across five validators.  So semuxpool_com generates 45 SEM a day and semux.info 225 SEM a day.

My 1400 votes on semux.info should pull in about 0.907 SEM/day (but they only payout weekly) based on (1400/347239)*225.  Or 6.35 SEM/week.

If I were to move my 1400 votes to semuxpool_com, assuming all else stayed equal, I should pull in (1400/(1400+62224))*45 = 0.990 SEM/day or 6.93 SEM/week.

However, to move my votes, because the semux.info pool is split across five validators, it would cost me 0.05*5 + 0.05 in transactions fees, or 0.30 SEM.  So, I'd make 0.08 SEM more per day on your pool, but it would take 4 days of earnings to pay for the transaction fees in changing my vote.  I guess I should leave my votes where they are until mainnet starts.

Is any part of my analysis faulty?

Oh... Do validators also get awarded transaction fees (like Bitcoin miners) or only a block reward?  That might change the calculus a little, but I think most semux blocks have been empty, so I assume there are very little transaction fees being spent now.