The effect you mentioned is real that when preferentially loading up bonds in tax deferred, it is sort-of equivalent to using an overall stock/bond allocation that is several points higher in stocks.
But I think eyenitnoy's regret was simply that had he been a better market timer, he could have done the conversions in the current market downturn, rather than at the peak. I did the same thing, figuring that on average doing my conversions early in the year would be better than waiting - oops! that didn't work this year.
I think it is very revealing that you mention that the results that eyenitnoy obtained depended on his
timing. I still contend that this is because he
changed his risk profile when he did the conversions. It is easy to lose sight of this, particularly when one does not consider funds in cash in a taxable account as part of one's portfolio.
To vastly oversimplify, he had some stock in a tIRA, and some cash in, say, a savings acount. He converted some of the stock to Roth, and paid for the taxes fom the cash. Clearly, he now has fewer
safe assets, and more (on an after-tax basis) risky ones.
To put some crude numbers, assume he had $400k in tIRA stocks, and $100k in taxable cash, for a total of nominal dollars of $500k. Let's say that his tax rate is 25%. Let's assume he coverted $200k in stocks.
Before the conversion, he had a
tax-adjusted portfolio of $400k*75% = $300k in stocks (all in tIRA), and $100k in cash, for a total of $400k in spendable, after-tax money.
After the conversion, he nominally had $200k in Roth stocks, $200k in tIRA stocks, and ($100k - 0.25*$200k)=$50k in taxable cash. His
tax-adjusted portfolio is now ($200k*75% tIRA stocks + $200k Roth stocks) = $350k in stocks; and cash reserves of $50k computed above, for a total of $400k spendable, after-tax money. This is as it should be: the conversion does not change his spendable, after-tax total.
However, note that his risk profile has changed. Before conversion, his tax-adjusted AA was something like 75/25 (stock/cash). After the conversion, his tax-adjusted AA was 87.5/12.5 (stock/cash). He has taken on more risk. A downturn affected him worse, and an updraft would have made him more happy.
He
could have mitigated this risk by maintaining the same tax-adjusted AA after converting that he deemed proper before converting. Then the timing would not have mattered.
The mistake being made here is to focus on individual parts of the portfolio, rather than the overall portfolio.