Here's how I see it.
First, if inflation is p (per week) then the corresponding deflation is q = -p/(1+p) [1].
Say a trader has production costs W and sells what is produced for R = 2W, each week (it's a little easier if we assume 100% mark up).
If now inflation kicks in at p, then on the first week W_1 = W(1+p) and R_1 = R(1+P), thus maintaining R/W = 2.
And on the kth week W_k = W(1+p)^k and R_k = R(1+p)^k, again maintaining R/W = 2.
The deflation case is trivially the same, with p replaced by q = -p/(1+p), and we find R/W = 2 as above.
That's it really.
You assume that goods are produced and sold instantaneously, which is not the case in real life. Production cycles can be as long as a few years. If the time span of your production cycle was equal to zero, then neither inflation nor deflation would have any impact on your profits (in percentages), which is what your example reveals.
Correct usage should be R_t2/W_t1, where t2 and t1 are different time moments for revenue and cost flows in a production cycle, t2 > t1. In inflation R_t2 is
always greater than W_t1 (provided we were profitable before inflation set in), whereas in deflation R_t2
may become less than W_t1 (even if we
were profitable before deflation set in, i.e. R > W and R/W time-invariant). That would mean a
loss. So, in inflation you can never mathematically suffer a loss due to inflation per se (if you were profitable before, of course), while in deflation it becomes quite possible through the effect of deflation as such.