I don't have a firm opinion one-way or another on this particular topic, but as a forensic toxicologist with too many years of experience in controlled substance-testing both in civil (workplace) and criminal cases, I do have an opinion on their suggested approach here, so I'll say this much . . .
Traditionally, blood testing was reserved for instances where 'time-specific' altered- or impaired- behavior of the subject was the issue at-hand (e.g. DUI/DWI in a vehicular manslaughter or other negligent homicide, unattended death/suicide investigations or coroners inquest, etc.). In most other cases however, breath analysis, urinalysis, even hair and saliva analyses proved convenient, expeditious and sufficient in protecting worker or public safety. That all changed with the advent of steroid abuse. Not only are steroids (and their key metabolites) difficult to detect by traditional means, they are also therapeutic (pharmacologically active) at extremely low concentrations in blood or urine, relative to the typical drug of abuse (e.g., cocaine, methamphetamine, etc.). This makes them a somewhat problematic group of analytes, and the U.S Olympic Committee learned that firsthand back in the late 80's.
Personally, if I'm the high-commissioner of golf, I don't much care about when a given subject was exposed to the controlled substance or illicit drug (be it anabolic steroid or otherwise), unless his/her actions were somehow 'time-sensitive' in nature, and let's face it, golf really ain't that. Some, like flamboyant boxer Floyd Mayweather, Jr., might argue that boxing matches or golf tournaments are in-fact "time-sensitive", but in the target case of steroids, are they, really? We're not talking here about the incidental, fast-acting influence of stimulants or depressants like amphetamines, cocaine, or ethyl alcohol on one's momentary driving ability. Rather, we're talking about the prolonged, chronic, low-level use/abuse of steroids and their purported influence on athletic prowess, and those are two radically different animals altogether.
How long ago the exposure occurred, how long a given drug is detectable, and in what specific biological matrices (blood or urine), should really be the only controlling factors in deciding whether to mandate blood vs. urine testing, and this evaluation in-turn hinges largely on the principal of "half-life" (or how long the drug is detectable in practical terms). As a commissioner or regulator of an athletic competition, if steroids are my chief concern, all that I should really care about, is whether the competitor in-question has been exposed to the regulated substance in his or her recent (perhaps 72-hour) history. Arguably, the use of blood for detecting the presence or absence of the parent steroid serves little to no purpose in this instance (i.e., detection of key metabolites hours or days after exposure via urinalysis is sufficiently incriminating in my view).
By example, when injected intravenously, heroin rapidly metabolizes to 6-mono-acetyl morphine (6-MAM), which then metabolizes to morphine and codeine, and those in-turn to still other (or further) opiates, and so-on. But the actual half-life of heroin itself (the parent drug) in whole blood, is typically less than 50-seconds. In other words, less than 1/100th of the initial dosage is present just ten-minutes later. So what good does it do me to test for it in blood? Or for that matter 6-MAM? They're both too short-lived! Unless the alleged infraction or crime was less than a couple of minutes ago, I've got no real chance of ever seeing it. Since the half-life of most drugs in blood (including steroids) is on the order of minutes or hours (not days or weeks), and since primary metabolites (in most cases) are more readily detected than the parent drug itself, its a non-starter. The parent drug continues to metabolize, and forensically significant metabolites continue to enter and pool-up (becoming more concentrated) in the individual's bladder for periods of up to 72-hours. So I say . . . who needs blood? Shoots, give me urine! Heroin or cocaine, androgens or estrogens, one-time abuser or frequent flyer, I''ll get it done perfectly well with pee-pee, baby! However, in the event that my preliminary results yield a confirmed positive in your urine, I'm likely to take a second run at you for a more invasive blood sample, and I'd then have probable cause to do.
Anyway, that's my 2-cents, but you didn't hear any of this from me! And if you try to say so, I'll deny it, most vehemently!